US20170199961A1 - Neoantigen Identification, Manufacture, and Use - Google Patents

Neoantigen Identification, Manufacture, and Use Download PDF

Info

Publication number
US20170199961A1
US20170199961A1 US15/381,375 US201615381375A US2017199961A1 US 20170199961 A1 US20170199961 A1 US 20170199961A1 US 201615381375 A US201615381375 A US 201615381375A US 2017199961 A1 US2017199961 A1 US 2017199961A1
Authority
US
United States
Prior art keywords
peptide
hla
allele
mhc
tumor
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Abandoned
Application number
US15/381,375
Other languages
English (en)
Inventor
Roman Yelensky
Adnan Derti
Brendan Bulik-Sullivan
Jennifer Busby
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Gritstone Bio Inc
Original Assignee
Gritstone Oncology Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Gritstone Oncology Inc filed Critical Gritstone Oncology Inc
Priority to US15/381,375 priority Critical patent/US20170199961A1/en
Priority to US15/466,729 priority patent/US10055540B2/en
Assigned to GRITSTONE ONCOLOGY, INC. reassignment GRITSTONE ONCOLOGY, INC. ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: BULIK-SULLIVAN, Brendan, BUSBY, Jennifer, DERTI, ADNAN, YELENSKY, Roman
Publication of US20170199961A1 publication Critical patent/US20170199961A1/en
Priority to US16/001,569 priority patent/US10847252B2/en
Priority to US16/040,409 priority patent/US10847253B2/en
Priority to US16/128,421 priority patent/US20190034585A1/en
Priority to US17/101,518 priority patent/US11183286B2/en
Priority to US17/101,522 priority patent/US20210166784A1/en
Assigned to GRITSTONE BIO, INC. reassignment GRITSTONE BIO, INC. CHANGE OF NAME (SEE DOCUMENT FOR DETAILS). Assignors: GRITSTONE ONCOLOGY, INC.
Abandoned legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
    • G16H20/10—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to drugs or medications, e.g. for ensuring correct administration to patients
    • G06F19/22—
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61K—PREPARATIONS FOR MEDICAL, DENTAL OR TOILETRY PURPOSES
    • A61K39/00—Medicinal preparations containing antigens or antibodies
    • A61K39/0005—Vertebrate antigens
    • A61K39/0011—Cancer antigens
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61K—PREPARATIONS FOR MEDICAL, DENTAL OR TOILETRY PURPOSES
    • A61K39/00—Medicinal preparations containing antigens or antibodies
    • A61K39/39—Medicinal preparations containing antigens or antibodies characterised by the immunostimulating additives, e.g. chemical adjuvants
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61P—SPECIFIC THERAPEUTIC ACTIVITY OF CHEMICAL COMPOUNDS OR MEDICINAL PREPARATIONS
    • A61P35/00—Antineoplastic agents
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61P—SPECIFIC THERAPEUTIC ACTIVITY OF CHEMICAL COMPOUNDS OR MEDICINAL PREPARATIONS
    • A61P35/00—Antineoplastic agents
    • A61P35/02—Antineoplastic agents specific for leukemia
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
    • G01N33/48—Biological material, e.g. blood, urine; Haemocytometers
    • G01N33/50—Chemical analysis of biological material, e.g. blood, urine; Testing involving biospecific ligand binding methods; Immunological testing
    • G01N33/53—Immunoassay; Biospecific binding assay; Materials therefor
    • G01N33/574—Immunoassay; Biospecific binding assay; Materials therefor for cancer
    • G01N33/57484—Immunoassay; Biospecific binding assay; Materials therefor for cancer involving compounds serving as markers for tumor, cancer, neoplasia, e.g. cellular determinants, receptors, heat shock/stress proteins, A-protein, oligosaccharides, metabolites
    • G—PHYSICS
    • G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
    • G09F—DISPLAYING; ADVERTISING; SIGNS; LABELS OR NAME-PLATES; SEALS
    • G09F19/00—Advertising or display means not otherwise provided for
    • G09F19/12—Advertising or display means not otherwise provided for using special optical effects
    • G09F19/16—Advertising or display means not otherwise provided for using special optical effects involving the use of mirrors
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
    • G16B20/00—ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
    • G16B30/00—ICT specially adapted for sequence analysis involving nucleotides or amino acids
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
    • G16B40/00—ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding
    • G16B40/10—Signal processing, e.g. from mass spectrometry [MS] or from PCR
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61K—PREPARATIONS FOR MEDICAL, DENTAL OR TOILETRY PURPOSES
    • A61K39/00—Medicinal preparations containing antigens or antibodies
    • A61K2039/51—Medicinal preparations containing antigens or antibodies comprising whole cells, viruses or DNA/RNA
    • A61K2039/515—Animal cells
    • A61K2039/5152—Tumor cells
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61K—PREPARATIONS FOR MEDICAL, DENTAL OR TOILETRY PURPOSES
    • A61K39/00—Medicinal preparations containing antigens or antibodies
    • A61K2039/51—Medicinal preparations containing antigens or antibodies comprising whole cells, viruses or DNA/RNA
    • A61K2039/515—Animal cells
    • A61K2039/5154—Antigen presenting cells [APCs], e.g. dendritic cells or macrophages
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61K—PREPARATIONS FOR MEDICAL, DENTAL OR TOILETRY PURPOSES
    • A61K39/00—Medicinal preparations containing antigens or antibodies
    • A61K2039/51—Medicinal preparations containing antigens or antibodies comprising whole cells, viruses or DNA/RNA
    • A61K2039/53—DNA (RNA) vaccination
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61K—PREPARATIONS FOR MEDICAL, DENTAL OR TOILETRY PURPOSES
    • A61K39/00—Medicinal preparations containing antigens or antibodies
    • A61K2039/57—Medicinal preparations containing antigens or antibodies characterised by the type of response, e.g. Th1, Th2
    • A61K2039/577—Medicinal preparations containing antigens or antibodies characterised by the type of response, e.g. Th1, Th2 tolerising response
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61K—PREPARATIONS FOR MEDICAL, DENTAL OR TOILETRY PURPOSES
    • A61K39/00—Medicinal preparations containing antigens or antibodies
    • A61K2039/58—Medicinal preparations containing antigens or antibodies raising an immune response against a target which is not the antigen used for immunisation
    • A61K2039/585—Medicinal preparations containing antigens or antibodies raising an immune response against a target which is not the antigen used for immunisation wherein the target is cancer
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61K—PREPARATIONS FOR MEDICAL, DENTAL OR TOILETRY PURPOSES
    • A61K35/00—Medicinal preparations containing materials or reaction products thereof with undetermined constitution
    • A61K35/12—Materials from mammals; Compositions comprising non-specified tissues or cells; Compositions comprising non-embryonic stem cells; Genetically modified cells
    • A61K35/14—Blood; Artificial blood
    • A61K35/15—Cells of the myeloid line, e.g. granulocytes, basophils, eosinophils, neutrophils, leucocytes, monocytes, macrophages or mast cells; Myeloid precursor cells; Antigen-presenting cells, e.g. dendritic cells
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N2333/00—Assays involving biological materials from specific organisms or of a specific nature
    • G01N2333/435—Assays involving biological materials from specific organisms or of a specific nature from animals; from humans
    • G01N2333/705—Assays involving receptors, cell surface antigens or cell surface determinants
    • G01N2333/70503—Immunoglobulin superfamily, e.g. VCAMs, PECAM, LFA-3
    • G01N2333/70539—MHC-molecules, e.g. HLA-molecules
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
    • G01N33/48—Biological material, e.g. blood, urine; Haemocytometers
    • G01N33/50—Chemical analysis of biological material, e.g. blood, urine; Testing involving biospecific ligand binding methods; Immunological testing
    • G01N33/68—Chemical analysis of biological material, e.g. blood, urine; Testing involving biospecific ligand binding methods; Immunological testing involving proteins, peptides or amino acids
    • G01N33/6803—General methods of protein analysis not limited to specific proteins or families of proteins
    • G01N33/6848—Methods of protein analysis involving mass spectrometry
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Definitions

  • neoantigen vaccine design is which of the many coding mutations present in subject tumors can generate the “best” therapeutic neoantigens, e.g., antigens that can elicit anti-tumor immunity and cause tumor regression.
  • previous approaches generated candidate neoantigens using only cis-acting mutations, and largely neglected to consider additional sources of neo-ORFs, including mutations in splicing factors, which occur in multiple tumor types and lead to aberrant splicing of many genes 13 , and mutations that create or remove protease cleavage sites.
  • standard approaches to tumor genome and transcriptome analysis can miss somatic mutations that give rise to candidate neoantigens due to suboptimal conditions in library construction, exome and transcriptome capture, sequencing, or data analysis.
  • standard tumor analysis approaches can inadvertently promote sequence artifacts or germline polymorphisms as neoantigens, leading to inefficient use of vaccine capacity or auto-immunity risk, respectively.
  • neoantigen candidate identification using next-generation sequencing (NGS) are addressed. These methods build on standard approaches for NGS tumor analysis to ensure that the highest sensitivity and specificity neoantigen candidates are advanced, across all classes of genomic alteration.
  • NGS next-generation sequencing
  • novel approaches for high-PPV neoantigen selection are presented to overcome the specificity problem and ensure that neoantigens advanced for vaccine inclusion are more likely to elicit anti-tumor immunity.
  • FIG. 1A shows current clinical approaches to neoantigen identification.
  • FIG. 1B shows that ⁇ 5% of predicted bound peptides are presented on tumor cells.
  • FIG. 1C shows the impact of the neoantigen prediction specificity problem.
  • FIG. 1D shows that binding prediction is not sufficient for neoantigen identification.
  • FIG. 1E shows probability of MHC-I presentation as a function of peptide length.
  • FIG. 1F shows an example peptide spectrum generated from Promega's dynamic range standard.
  • Figure discloses SEQ ID NO: 1.
  • FIG. 1G shows how the addition of features increases the model positive predictive value.
  • FIG. 2A is an overview of an environment for identifying likelihoods of peptide presentation in patients, in accordance with an embodiment.
  • FIGS. 2B and 2C illustrate a method of obtaining presentation information, in accordance with an embodiment.
  • FIG. 2B discloses SEQ ID NO: 3.
  • FIG. 2C discloses SEQ ID NOS 3-8, respectively, in order of appearance.
  • FIG. 3 is a high-level block diagram illustrating the computer logic components of the presentation identification system, according to one embodiment.
  • FIG. 4 illustrates an example set of training data, according to one embodiment.
  • Figure discloses the “Peptide Sequences” as SEQ ID NOS 10-13 and the “C-Flanking Sequences” as SEQ ID NOS 14, 19-20, and 20, respectively, in order of appearance.
  • FIG. 5 illustrates an example network model in association with an MHC allele.
  • FIG. 6A illustrates an example network model NN H ( ⁇ ) shared by MHC alleles, according to one embodiment.
  • FIG. 6B illustrates an example network model NN H ( ⁇ ) shared by MHC alleles, according to another embodiment.
  • FIG. 7 illustrates generating a presentation likelihood for a peptide in association with an MHC allele using an example network model.
  • FIG. 8 illustrates generating a presentation likelihood for a peptide in association with a MHC allele using example network models.
  • FIG. 9 illustrates generating a presentation likelihood for a peptide in association with MHC alleles using example network models.
  • FIG. 10 illustrates generating a presentation likelihood for a peptide in association with MHC alleles using example network models.
  • FIG. 11 illustrates generating a presentation likelihood for a peptide in association with MHC alleles using example network models.
  • FIG. 12 illustrates generating a presentation likelihood for a peptide in association with MHC alleles using example network models.
  • FIG. 13A compares performance results of an example presentation model and state-of-the-art models for predicting peptide presentation on multiple-allele mass spectrometry data.
  • FIG. 13B compares performance results of another example presentation model, and state-of-the-art models for predicting peptide presentation on T-cell epitope data.
  • FIG. 13C compares performance results for an example function-of-sums model, an example sum-of-functions model, and an example second order model for predicting peptide presentation on multiple-allele mass spectrometry data.
  • FIG. 13D compares performance results for two example presentation models that are trained with and without single-allele mass spectrometry data on predicting peptide presentation for multiple-allele mass spectrometry data.
  • FIG. 13E shows performance for example models shown in FIG. 13D on single-allele mass spectrometry data that were held out in the analysis shown in FIG. 13D .
  • FIG. 13F shows the common anchor residues at positions 2 and 9 among nonamers predicted by the “without A2/B7 single-allele data” example model shown in FIG. 13D .
  • FIG. 13G compares performance results between an example presentation model that incorporated C- and N-terminal flanking sequences as allele-interacting variables, and an example presentation model that incorporated C- and N-terminal flanking sequences as allele-noninteracting variables.
  • FIG. 13H illustrates the dependency between fraction of presented peptides for genes based on mRNA quantification for mass spectrometry data on tumor cells.
  • FIG. 13I shows performance of two example presentation models, one of which is trained based on mass spectrometry tumor cell data, another of which incorporates mRNA quantification data and mass spectrometry tumor cell data.
  • FIG. 14 illustrates an example computer for implementing the entities shown in FIGS. 1 and 3 .
  • the term “antigen” is a substance that induces an immune response.
  • neoantigen is an antigen that has at least one alteration that makes it distinct from the corresponding wild-type, parental antigen, e.g., via mutation in a tumor cell or post-translational modification specific to a tumor cell.
  • a neoantigen can include a polypeptide sequence or a nucleotide sequence.
  • a mutation can include a frameshift or nonframeshift indel, missense or nonsense substitution, splice site alteration, genomic rearrangement or gene fusion, or any genomic or expression alteration giving rise to a neoORF.
  • a mutations can also include a splice variant.
  • Post-translational modifications specific to a tumor cell can include aberrant phosphorylation.
  • Post-translational modifications specific to a tumor cell can also include a proteasome-generated spliced antigen. See Liepe et al., A large fraction of HLA class I ligands are proteasome-generated spliced peptides; Science. 2016 Oct. 21; 354(6310):354-358.
  • tumor neoantigen is a neoantigen present in a subject's tumor cell or tissue but not in the subject's corresponding normal cell or tissue.
  • neoantigen-based vaccine is a vaccine construct based on one or more neoantigens, e.g., a plurality of neoantigens.
  • candidate neoantigen is a mutation or other aberration giving rise to a new sequence that may represent a neoantigen.
  • coding region is the portion(s) of a gene that encode protein.
  • coding mutation is a mutation occurring in a coding region.
  • ORF means open reading frame
  • NEO-ORF is a tumor-specific ORF arising from a mutation or other aberration such as splicing.
  • missense mutation is a mutation causing a substitution from one amino acid to another.
  • nonsense mutation is a mutation causing a substitution from an amino acid to a stop codon.
  • frameshift mutation is a mutation causing a change in the frame of the protein.
  • the term “indel” is an insertion or deletion of one or more nucleic acids.
  • the term percent “identity,” in the context of two or more nucleic acid or polypeptide sequences, refer to two or more sequences or subsequences that have a specified percentage of nucleotides or amino acid residues that are the same, when compared and aligned for maximum correspondence, as measured using one of the sequence comparison algorithms described below (e.g., BLASTP and BLASTN or other algorithms available to persons of skill) or by visual inspection.
  • the percent “identity” can exist over a region of the sequence being compared, e.g., over a functional domain, or, alternatively, exist over the full length of the two sequences to be compared.
  • sequence comparison typically one sequence acts as a reference sequence to which test sequences are compared.
  • test and reference sequences are input into a computer, subsequence coordinates are designated, if necessary, and sequence algorithm program parameters are designated.
  • sequence comparison algorithm then calculates the percent sequence identity for the test sequence(s) relative to the reference sequence, based on the designated program parameters.
  • sequence similarity or dissimilarity can be established by the combined presence or absence of particular nucleotides, or, for translated sequences, amino acids at selected sequence positions (e.g., sequence motifs).
  • Optimal alignment of sequences for comparison can be conducted, e.g., by the local homology algorithm of Smith & Waterman, Adv. Appl. Math. 2:482 (1981), by the homology alignment algorithm of Needleman & Wunsch, J. Mol. Biol. 48:443 (1970), by the search for similarity method of Pearson & Lipman, Proc. Nat'l. Acad. Sci. USA 85:2444 (1988), by computerized implementations of these algorithms (GAP, BESTFIT, FASTA, and TFASTA in the Wisconsin Genetics Software Package, Genetics Computer Group, 575 Science Dr., Madison, Wis.), or by visual inspection (see generally Ausubel et al., infra).
  • BLAST algorithm One example of an algorithm that is suitable for determining percent sequence identity and sequence similarity is the BLAST algorithm, which is described in Altschul et al., J. Mol. Biol. 215:403-410 (1990). Software for performing BLAST analyses is publicly available through the National Center for Biotechnology Information.
  • non-stop or read-through is a mutation causing the removal of the natural stop codon.
  • epitopope is the specific portion of an antigen typically bound by an antibody or T cell receptor.
  • immunogenic is the ability to elicit an immune response, e.g., via T cells, B cells, or both.
  • HLA binding affinity means affinity of binding between a specific antigen and a specific MHC allele.
  • the term “bait” is a nucleic acid probe used to enrich a specific sequence of DNA or RNA from a sample.
  • variable is a difference between a subject's nucleic acids and the reference human genome used as a control.
  • variant call is an algorithmic determination of the presence of a variant, typically from sequencing.
  • polymorphism is a germline variant, i.e., a variant found in all DNA-bearing cells of an individual.
  • matic variant is a variant arising in non-germline cells of an individual.
  • allele is a version of a gene or a version of a genetic sequence or a version of a protein.
  • HLA type is the complement of HLA gene alleles.
  • nonsense-mediated decay or “NMD” is a degradation of an mRNA by a cell due to a premature stop codon.
  • truncal mutation is a mutation originating early in the development of a tumor and present in a substantial portion of the tumor's cells.
  • subclonal mutation is a mutation originating later in the development of a tumor and present in only a subset of the tumor's cells.
  • exome is a subset of the genome that codes for proteins.
  • An exome can be the collective exons of a genome.
  • logistic regression is a regression model for binary data from statistics where the logit of the probability that the dependent variable is equal to one is modeled as a linear function of the dependent variables.
  • neural network is a machine learning model for classification or regression consisting of multiple layers of linear transformations followed by element-wise nonlinearities typically trained via stochastic gradient descent and back-propagation.
  • proteome is the set of all proteins expressed and/or translated by a cell, group of cells, or individual.
  • peptidome is the set of all peptides presented by MHC-I or MHC-II on the cell surface.
  • the peptidome may refer to a property of a cell or a collection of cells (e.g., the tumor peptidome, meaning the union of the peptidomes of all cells that comprise the tumor).
  • ELISPOT Enzyme-linked immunosorbent spot assay—which is a common method for monitoring immune responses in humans and animals.
  • extracts is a dextran-based peptide-MHC multimers used for antigen-specific T-cell staining in flow cytometry.
  • tolerance or immune tolerance is a state of immune non-responsiveness to one or more antigens, e.g. self-antigens.
  • central tolerance is a tolerance affected in the thymus, either by deleting self-reactive T-cell clones or by promoting self-reactive T-cell clones to differentiate into immunosuppressive regulatory T-cells (Tregs).
  • peripheral tolerance is a tolerance affected in the periphery by downregulating or anergizing self-reactive T-cells that survive central tolerance or promoting these T cells to differentiate into Tregs.
  • sample can include a single cell or multiple cells or fragments of cells or an aliquot of body fluid, taken from a subject, by means including venipuncture, excretion, ejaculation, massage, biopsy, needle aspirate, lavage sample, scraping, surgical incision, or intervention or other means known in the art.
  • subject encompasses a cell, tissue, or organism, human or non-human, whether in vivo, ex vivo, or in vitro, male or female.
  • subject is inclusive of mammals including humans.
  • mammal encompasses both humans and non-humans and includes but is not limited to humans, non-human primates, canines, felines, murines, bovines, equines, and porcines.
  • Clinical factor refers to a measure of a condition of a subject, e.g., disease activity or severity.
  • “Clinical factor” encompasses all markers of a subject's health status, including non-sample markers, and/or other characteristics of a subject, such as, without limitation, age and gender.
  • a clinical factor can be a score, a value, or a set of values that can be obtained from evaluation of a sample (or population of samples) from a subject or a subject under a determined condition.
  • a clinical factor can also be predicted by markers and/or other parameters such as gene expression surrogates.
  • Clinical factors can include tumor type, tumor sub-type, and smoking history.
  • MHC major histocompatibility complex
  • HLA human leukocyte antigen, or the human MHC gene locus
  • NGS next-generation sequencing
  • PPV positive predictive value
  • TSNA tumor-specific neoantigen
  • FFPE formalin-fixed, paraffin-embedded
  • NMD nonsense-mediated decay
  • NSCLC non-small-cell lung cancer
  • DC dendritic cell.
  • one such method may comprise the steps of: obtaining at least one of exome, transcriptome or whole genome tumor nucleotide sequencing data from the tumor cell of the subject, wherein the tumor nucleotide sequencing data is used to obtain data representing peptide sequences of each of a set of neoantigens, and wherein the peptide sequence of each neoantigen comprises at least one alteration that makes it distinct from the corresponding wild-type, parental peptide sequence; inputting the peptide sequence of each neoantigen into one or more presentation models to generate a set of numerical likelihoods that each of the neoantigens is presented by one or more MHC alleles on the tumor cell surface of the tumor cell of the subject or cells present in the tumor, the set of numerical likelihoods having been identified at least based on received
  • the presentation model can comprise a statistical regression or a machine learning (e.g., deep learning) model trained on a set of reference data (also referred to as a training data set) comprising a set of corresponding labels, wherein the set of reference data is obtained from each of a plurality of distinct subjects where optionally some subjects can have a tumor, and wherein the set of reference data comprises at least one of: data representing exome nucleotide sequences from tumor tissue, data representing exome nucleotide sequences from normal tissue, data representing transcriptome nucleotide sequences from tumor tissue, data representing proteome sequences from tumor tissue, and data representing MHC peptidome sequences from tumor tissue, and data representing MHC peptidome sequences from normal tissue.
  • a machine learning e.g., deep learning
  • the reference data can further comprise mass spectrometry data, sequencing data, RNA sequencing data, and proteomics data for single-allele cell lines engineered to express a predetermined MHC allele that are subsequently exposed to synthetic protein, normal and tumor human cell lines, and fresh and frozen primary samples, and T cell assays (e.g., ELISPOT).
  • the set of reference data includes each form of reference data.
  • the presentation model can comprise a set of features derived at least in part from the set of reference data, and wherein the set of features comprises at least one of allele dependent-features and allele-independent features. In certain aspects each feature is included.
  • Dendritic cell presentation to na ⁇ ve T cell features can comprise at least one of: A feature described above.
  • the dose and type of antigen in the vaccine. e.g., peptide, mRNA, virus, etc.: (1) The route by which dendritic cells (DCs) take up the antigen type (e.g., endocytosis, micropinocytosis); and/or (2) The efficacy with which the antigen is taken up by DCs.
  • the dose and type of adjuvant in the vaccine The length of the vaccine antigen sequence. The number and sites of vaccine administration. Baseline patient immune functioning (e.g., as measured by history of recent infections, blood counts, etc).
  • RNA vaccines (1) the turnover rate of the mRNA protein product in the dendritic cell; (2) the rate of translation of the mRNA after uptake by dendritic cells as measured in in vitro or in vivo experiments; and/or (3) the number or rounds of translation of the mRNA after uptake by dendritic cells as measured by in vivo or in vitro experiments.
  • the level of expression of the proteasome and immunoproteasome in typical activated dendritic cells (which may be measured by RNA-seq, mass spectrometry, immunohistochemistry, or other standard techniques).
  • the expression levels of the particular MHC allele in the individual in question e.g., as measured by RNA-seq or mass spectrometry, optionally measured specifically in activated dendritic cells or other immune cells.
  • the probability of peptide presentation by the particular MHC allele in other individuals who express the particular MHC allele optionally measured specifically in activated dendritic cells or other immune cells.
  • the probability of peptide presentation by MHC alleles in the same family of molecules e.g., HLA-A, HLA-B, HLA-C, HLA-DQ, HLA-DR, HLA-DP
  • HLA-A, HLA-B, HLA-C, HLA-DQ, HLA-DR, HLA-DP optionally measured specifically in activated dendritic cells or other immune cells.
  • Immune tolerance escape features can comprise at least one of: Direct measurement of the self-peptidome via protein mass spectrometry performed on one or several cell types. Estimation of the self-peptidome by taking the union of all k-mer (e.g. 5-25) substrings of self-proteins. Estimation of the self-peptidome using a model of presentation similar to the presentation model described above applied to all non-mutation self-proteins, optionally accounting for germline variants.
  • Ranking can be performed using the plurality of neoantigens provided by at least one model based at least in part on the numerical likelihoods. Following the ranking a selecting can be performed to select a subset of the ranked neoantigens according to a selection criteria. After selecting a subset of the ranked peptides can be provided as an output.
  • a number of the set of selected neoantigens may be 20.
  • the presentation model may represent dependence between presence of a pair of a particular one of the MHC alleles and a particular amino acid at a particular position of a peptide sequence; and likelihood of presentation on the tumor cell surface, by the particular one of the MHC alleles of the pair, of such a peptide sequence comprising the particular amino acid at the particular position.
  • a method disclosed herein can also include applying the one or more presentation models to the peptide sequence of the corresponding neoantigen to generate a dependency score for each of the one or more MHC alleles indicating whether the MHC allele will present the corresponding neoantigen based on at least positions of amino acids of the peptide sequence of the corresponding neoantigen.
  • a method disclosed herein can also include transforming the dependency scores to generate a corresponding per-allele likelihood for each MHC allele indicating a likelihood that the corresponding MHC allele will present the corresponding neoantigen; and combining the per-allele likelihoods to generate the numerical likelihood.
  • the step of transforming the dependency scores can model the presentation of the peptide sequence of the corresponding neoantigen as mutually exclusive.
  • a method disclosed herein can also include transforming a combination of the dependency scores to generate the numerical likelihood.
  • the step of transforming the combination of the dependency scores can model the presentation of the peptide sequence of the corresponding neoantigen as interfering between MHC alleles.
  • the set of numerical likelihoods can be further identified by at least an allele noninteracting feature, and a method disclosed herein can also include applying an allele noninteracting one of the one or more presentation models to the allele noninteracting features to generate a dependency score for the allele noninteracting features indicating whether the peptide sequence of the corresponding neoantigen will be presented based on the allele noninteracting features.
  • a method disclosed herein can also include combining the dependency score for each MHC allele in the one or more MHC alleles with the dependency score for the allele noninteracting feature; transforming the combined dependency scores for each MHC allele to generate a corresponding per-allele likelihood for the MHC allele indicating a likelihood that the corresponding MHC allele will present the corresponding neoantigen; and combining the per-allele likelihoods to generate the numerical likelihood.
  • a method disclosed herein can also include transforming a combination of the dependency scores for each of the MHC alleles and the dependency score for the allele noninteracting features to generate the numerical likelihood.
  • a set of numerical parameters for the presentation model can be trained based on a training data set including at least a set of training peptide sequences identified as present in a plurality of samples and one or more MHC alleles associated with each training peptide sequence, wherein the training peptide sequences are identified through mass spectrometry on isolated peptides eluted from MHC alleles derived from the plurality of samples.
  • the samples can also include cell lines engineered to express a single MHC class I or class II allele.
  • the samples can also include cell lines engineered to express a plurality of MHC class I or class II alleles.
  • the samples can also include human cell lines obtained or derived from a plurality of patients.
  • the samples can also include fresh or frozen tumor samples obtained from a plurality of patients.
  • the samples can also include fresh or frozen tissue samples obtained from a plurality of patients.
  • the samples can also include peptides identified using T-cell assays.
  • the training data set can further include data associated with: peptide abundance of the set of training peptides present in the samples; peptide length of the set of training peptides in the samples.
  • the training data set may be generated by comparing the set of training peptide sequences via alignment to a database comprising a set of known protein sequences, wherein the set of training protein sequences are longer than and include the training peptide sequences.
  • the training data set may be generated based on performing or having performed nucleotide sequencing on a cell line to obtain at least one of exome, transcriptome, or whole genome sequencing data from the cell line, the sequencing data including at least one nucleotide sequence including an alteration.
  • the training data set may be generated based on obtaining at least one of exome, transcriptome, and whole genome normal nucleotide sequencing data from normal tissue samples.
  • the training data set may further include data associated with proteome sequences associated with the samples.
  • the training data set may further include data associated with MHC peptidome sequences associated with the samples.
  • the training data set may further include data associated with peptide-MHC binding affinity measurements for at least one of the isolated peptides.
  • the training data set may further include data associated with peptide-MHC binding stability measurements for at least one of the isolated peptides.
  • the training data set may further include data associated with transcriptomes associated with the samples.
  • the training data set may further include data associated with genomes associated with the samples.
  • the training peptide sequences may be of lengths within a range of k-mers where k is between 8-15, inclusive for MHC class I or 9-30 inclusive for MHC class II.
  • a method disclosed herein can also include encoding the peptide sequence using a one-hot encoding scheme.
  • a method disclosed herein can also include encoding the training peptide sequences using a left-padded one-hot encoding scheme.
  • a method of treating a subject having a tumor comprising performing the steps of claim 1 , and further comprising obtaining a tumor vaccine comprising the set of selected neoantigens, and administering the tumor vaccine to the subject.
  • Also disclosed herein is a methods for manufacturing a tumor vaccine, comprising the steps of: obtaining at least one of exome, transcriptome or whole genome tumor nucleotide sequencing data from the tumor cell of the subject, wherein the tumor nucleotide sequencing data is used to obtain data representing peptide sequences of each of a set of neoantigens, and wherein the peptide sequence of each neoantigen comprises at least one mutation that makes it distinct from the corresponding wild-type, parental peptide sequence; inputting the peptide sequence of each neoantigen into one or more presentation models to generate a set of numerical likelihoods that each of the neoantigens is presented by one or more MHC alleles on the tumor cell surface of the tumor cell of the subject, the set of numerical likelihoods having been identified at least based on received mass spectrometry data; and selecting a subset of the set of neoantigens based on the set of numerical likelihoods to generate a set of selected neoantigens
  • a tumor vaccine including a set of selected neoantigens selected by performing the method comprising the steps of: obtaining at least one of exome, transcriptome or whole genome tumor nucleotide sequencing data from the tumor cell of the subject, wherein the tumor nucleotide sequencing data is used to obtain data representing peptide sequences of each of a set of neoantigens, and wherein the peptide sequence of each neoantigen comprises at least one mutation that makes it distinct from the corresponding wild-type, parental peptide sequence; inputting the peptide sequence of each neoantigen into one or more presentation models to generate a set of numerical likelihoods that each of the neoantigens is presented by one or more MHC alleles on the tumor cell surface of the tumor cell of the subject, the set of numerical likelihoods having been identified at least based on received mass spectrometry data; and selecting a subset of the set of neoantigens based on the set of numerical likelihoods to generate
  • the tumor vaccine may include one or more of a nucleotide sequence, a polypeptide sequence, RNA, DNA, a cell, a plasmid, or a vector.
  • the tumor vaccine may include one or more neoantigens presented on the tumor cell surface.
  • the tumor vaccine may include one or more neoantigens that is immunogenic in the subject.
  • the tumor vaccine may not include one or more neoantigens that induce an autoimmune response against normal tissue in the subject.
  • the tumor vaccine may include an adjuvant.
  • the tumor vaccine may include an excipient.
  • a method disclosed herein may also include selecting neoantigens that have an increased likelihood of being presented on the tumor cell surface relative to unselected neoantigens based on the presentation model.
  • a method disclosed herein may also include selecting neoantigens that have an increased likelihood of being capable of inducing a tumor-specific immune response in the subject relative to unselected neoantigens based on the presentation model.
  • a method disclosed herein may also include selecting neoantigens that have an increased likelihood of being capable of being presented to na ⁇ ve T cells by professional antigen presenting cells (APCs) relative to unselected neoantigens based on the presentation model, optionally wherein the APC is a dendritic cell (DC).
  • APCs professional antigen presenting cells
  • DC dendritic cell
  • a method disclosed herein may also include selecting neoantigens that have a decreased likelihood of being subject to inhibition via central or peripheral tolerance relative to unselected neoantigens based on the presentation model.
  • a method disclosed herein may also include selecting neoantigens that have a decreased likelihood of being capable of inducing an autoimmune response to normal tissue in the subject relative to unselected neoantigens based on the presentation model.
  • the exome or transcriptome nucleotide sequencing data may be obtained by performing sequencing on the tumor tissue.
  • the sequencing may be next generation sequencing (NGS) or any massively parallel sequencing approach.
  • NGS next generation sequencing
  • massively parallel sequencing approach any massively parallel sequencing approach.
  • the set of numerical likelihoods may be further identified by at least MHC-allele interacting features comprising at least one of: the predicted affinity with which the MHC allele and the neoantigen encoded peptide bind; the predicted stability of the neoantigen encoded peptide-MHC complex; the sequence and length of the neoantigen encoded peptide; the probability of presentation of neoantigen encoded peptides with similar sequence in cells from other individuals expressing the particular MHC allele as assessed by mass-spectrometry proteomics or other means; the expression levels of the particular MHC allele in the subject in question (e.g.
  • RNA-seq or mass spectrometry the overall neoantigen encoded peptide-sequence-independent probability of presentation by the particular MHC allele in other distinct subjects who express the particular MHC allele; the overall neoantigen encoded peptide-sequence-independent probability of presentation by MHC alleles in the same family of molecules (e.g., HLA-A, HLA-B, HLA-C, HLA-DQ, HLA-DR, HLA-DP) in other distinct subjects.
  • HLA-A, HLA-B, HLA-C, HLA-DQ, HLA-DR, HLA-DP in other distinct subjects.
  • the set of numerical likelihoods are further identified by at least MHC-allele noninteracting features comprising at least one of: the C- and N-terminal sequences flanking the neoantigen encoded peptide within its source protein sequence; the presence of protease cleavage motifs in the neoantigen encoded peptide, optionally weighted according to the expression of corresponding proteases in the tumor cells (as measured by RNA-seq or mass spectrometry); the turnover rate of the source protein as measured in the appropriate cell type; the length of the source protein, optionally considering the specific splice variants (“isoforms”) most highly expressed in the tumor cells as measured by RNA-seq or proteome mass spectrometry, or as predicted from the annotation of germline or somatic splicing mutations detected in DNA or RNA sequence data; the level of expression of the proteasome, immunoproteasome, thymoproteasome, or other proteases in the tumor cells (which may be measured by
  • Peptides whose presentation relies on a component of the antigen-presentation machinery that is subject to loss-of-function mutation in the tumor have reduced probability of presentation; presence or absence of functional germline polymorphisms, including, but not limited to: in genes encoding the proteins involved in the antigen presentation machinery (e.g., B2M, HLA-A, HLA-B, HLA-C, TAP-1, TAP-2, TAPBP, CALR, CNX, ERP57, HLA-DM, HLA-DMA, HLA-DMB, HLA-DO, HLA-DOA, HLA-DOBHLA-DP, HLA-DPA1, HLA-DPB1, HLA-DQ, HLA-DQA1, HLA-DQA2, HLA-DQB1, HLA-DQB2, HLA-DR, HLA-DRA, HLA-DRB1, HLA-DRB3, HLA-DRB4, HLA-DRB5 or any of the genes coding for components of the proteasome or immunoprote
  • the at least one mutation may be a frameshift or nonframeshift indel, missense or nonsense substitution, splice site alteration, genomic rearrangement or gene fusion, or any genomic or expression alteration giving rise to a neoORF.
  • the tumor cell may be selected from the group consisting of: lung cancer, melanoma, breast cancer, ovarian cancer, prostate cancer, kidney cancer, gastric cancer, colon cancer, testicular cancer, head and neck cancer, pancreatic cancer, brain cancer, B-cell lymphoma, acute myelogenous leukemia, chronic myelogenous leukemia, chronic lymphocytic leukemia, and T cell lymphocytic leukemia, non-small cell lung cancer, and small cell lung cancer.
  • a method disclosed herein may also include obtaining a tumor vaccine comprising the set of selected neoantigens or a subset thereof, optionally further comprising administering the tumor vaccine to the subject.
  • At least one of neoantigens in the set of selected neoantigens, when in polypeptide form, may include at least one of: a binding affinity with MHC with an IC50 value of less than 1000 nM, for MHC Class 1 polypeptides a length of 8-15, 8, 9, 10, 11, 12, 13, 14, or 15 amino acids, presence of sequence motifs within or near the polypeptide in the parent protein sequence promoting proteasome cleavage, and presence of sequence motifs promoting TAP transport.
  • Also disclosed herein is a methods for generating a model for identifying one or more neoantigens that are likely to be presented on a tumor cell surface of a tumor cell, comprising the steps of: receiving mass spectrometry data comprising data associated with a plurality of isolated peptides eluted from major histocompatibility complex (MHC) derived from a plurality of samples; obtaining a training data set by at least identifying a set of training peptide sequences present in the samples and one or more MHCs associated with each training peptide sequence; training a set of numerical parameters of a presentation model using the training data set comprising the training peptide sequences, the presentation model providing a plurality of numerical likelihoods that peptide sequences from the tumor cell are presented by one or more MHC alleles on the tumor cell surface.
  • MHC major histocompatibility complex
  • the presentation model may represent dependence between: presence of a particular amino acid at a particular position of a peptide sequence; and likelihood of presentation, by one of the MHC alleles on the tumor cell, of the peptide sequence containing the particular amino acid at the particular position.
  • the samples can also include cell lines engineered to express a single MHC class I or class II allele.
  • the samples can also include cell lines engineered to express a plurality of MHC class I or class II alleles.
  • the samples can also include human cell lines obtained or derived from a plurality of patients.
  • the samples can also include fresh or frozen tumor samples obtained from a plurality of patients.
  • the samples can also include peptides identified using T-cell assays.
  • the training data set may further include data associated with: peptide abundance of the set of training peptides present in the samples; peptide length of the set of training peptides in the samples.
  • a method disclosed herein can also include obtaining a set of training protein sequences based on the training peptide sequences by comparing the set of training peptide sequences via alignment to a database comprising a set of known protein sequences, wherein the set of training protein sequences are longer than and include the training peptide sequences.
  • a method disclosed herein can also include performing or having performed mass spectrometry on a cell line to obtain at least one of exome, transcriptome, or whole genome nucleotide sequencing data from the cell line, the nucleotide sequencing data including at least one protein sequence including a mutation.
  • a method disclosed herein can also include: encoding the training peptide sequences using a one-hot encoding scheme.
  • a method disclosed herein can also include obtaining at least one of exome, transcriptome, and whole genome normal nucleotide sequencing data from normal tissue samples; and training the set of parameters of the presentation model using the normal nucleotide sequencing data.
  • the training data set may further include data associated with proteome sequences associated with the samples.
  • the training data set may further include data associated with MHC peptidome sequences associated with the samples.
  • the training data set may further include data associated with peptide-MHC binding affinity measurements for at least one of the isolated peptides.
  • the training data set may further include data associated with peptide-MHC binding stability measurements for at least one of the isolated peptides.
  • the training data set may further include data associated with transcriptomes associated with the samples.
  • the training data set may further include data associated with genomes associated with the samples.
  • a method disclosed herein may also include logistically regressing the set of parameters.
  • the training peptide sequences may be lengths within a range of k-mers where k is between 8-15, inclusive for MHC class I or 9-30, inclusive for MHC class II.
  • a method disclosed herein may also include encoding the training peptide sequences using a left-padded one-hot encoding scheme.
  • a method disclosed herein may also include determining values for the set of parameters using a deep learning algorithm.
  • identifying one or more neoantigens that are likely to be presented on a tumor cell surface of a tumor cell comprising executing the steps of: receiving mass spectrometry data comprising data associated with a plurality of isolated peptides eluted from major histocompatibility complex (MHC) derived from a plurality of fresh or frozen tumor samples; obtaining a training data set by at least identifying a set of training peptide sequences present in the tumor samples and presented on one or more MHC alleles associated with each training peptide sequence; obtaining a set of training protein sequences based on the training peptide sequences; and training a set of numerical parameters of a presentation model using the training protein sequences and the training peptide sequences, the presentation model providing a plurality of numerical likelihoods that peptide sequences from the tumor cell are presented by one or more MHC alleles on the tumor cell surface.
  • MHC major histocompatibility complex
  • the presentation model may represent dependence between: presence of a pair of a particular one of the MHC alleles and a particular amino acid at a particular position of a peptide sequence; and likelihood of presentation on the tumor cell surface, by the particular one of the MHC alleles of the pair, of such a peptide sequence comprising the particular amino acid at the particular position.
  • a method disclosed herein can also include selecting a subset of neoantigens, wherein the subset of neoantigens is selected because each has an increased likelihood that it is presented on the cell surface of the tumor relative to one or more distinct tumor neoantigens.
  • a method disclosed herein can also include selecting a subset of neoantigens, wherein the subset of neoantigens is selected because each has an increased likelihood that it is capable of inducing a tumor-specific immune response in the subject relative to one or more distinct tumor neoantigens.
  • a method disclosed herein can also include selecting a subset of neoantigens, wherein the subset of neoantigens is selected because each has an increased likelihood that it is capable of being presented to na ⁇ ve T cells by professional antigen presenting cells (APCs) relative to one or more distinct tumor neoantigens, optionally wherein the APC is a dendritic cell (DC).
  • APCs professional antigen presenting cells
  • DC dendritic cell
  • a method disclosed herein can also include selecting a subset of neoantigens, wherein the subset of neoantigens is selected because each has a decreased likelihood that it is subject to inhibition via central or peripheral tolerance relative to one or more distinct tumor neoantigens.
  • a method disclosed herein can also include selecting a subset of neoantigens, wherein the subset of neoantigens is selected because each has a decreased likelihood that it is capable of inducing an autoimmune response to normal tissue in the subject relative to one or more distinct tumor neoantigens.
  • a method disclosed herein can also include selecting a subset of neoantigens, wherein the subset of neoantigens is selected because each has a decreased likelihood that it will be differentially post-translationally modified in tumor cells versus APCs, optionally wherein the APC is a dendritic cell (DC).
  • DC dendritic cell
  • mutations e.g., the variants or alleles that are present in cancer cells.
  • these mutations can be present in the genome, transcriptome, proteome, or exome of cancer cells of a subject having cancer but not in normal tissue from the subject.
  • Genetic mutations in tumors can be considered useful for the immunological targeting of tumors if they lead to changes in the amino acid sequence of a protein exclusively in the tumor.
  • Useful mutations include: (1) non-synonymous mutations leading to different amino acids in the protein; (2) read-through mutations in which a stop codon is modified or deleted, leading to translation of a longer protein with a novel tumor-specific sequence at the C-terminus; (3) splice site mutations that lead to the inclusion of an intron in the mature mRNA and thus a unique tumor-specific protein sequence; (4) chromosomal rearrangements that give rise to a chimeric protein with tumor-specific sequences at the junction of 2 proteins (i.e., gene fusion); (5) frameshift mutations or deletions that lead to a new open reading frame with a novel tumor-specific protein sequence. Mutations can also include one or more of nonframeshift indel, missense or nonsense substitution, splice site alteration, genomic rearrangement or gene fusion, or any genomic or
  • Peptides with mutations or mutated polypeptides arising from for example, splice-site, frameshift, readthrough, or gene fusion mutations in tumor cells can be identified by sequencing DNA, RNA or protein in tumor versus normal cells.
  • mutations can include previously identified tumor specific mutations. Known tumor mutations can be found at the Catalogue of Somatic Mutations in Cancer (COSMIC) database.
  • DASH dynamic allele-specific hybridization
  • MADGE microplate array diagonal gel electrophoresis
  • pyrosequencing oligonucleotide-specific ligation
  • PCR based detection means can include multiplex amplification of a plurality of markers simultaneously. For example, it is well known in the art to select PCR primers to generate PCR products that do not overlap in size and can be analyzed simultaneously. Alternatively, it is possible to amplify different markers with primers that are differentially labeled and thus can each be differentially detected. Of course, hybridization based detection means allow the differential detection of multiple PCR products in a sample. Other techniques are known in the art to allow multiplex analyses of a plurality of markers.
  • RNA molecules obtained from genomic DNA or cellular RNA.
  • a single base polymorphism can be detected by using a specialized exonuclease-resistant nucleotide, as disclosed, e.g., in Mundy, C. R. (U.S. Pat. No. 4,656,127).
  • a primer complementary to the allelic sequence immediately 3′ to the polymorphic site is permitted to hybridize to a target molecule obtained from a particular animal or human.
  • the polymorphic site on the target molecule contains a nucleotide that is complementary to the particular exonuclease-resistant nucleotide derivative present, then that derivative will be incorporated onto the end of the hybridized primer. Such incorporation renders the primer resistant to exonuclease, and thereby permits its detection. Since the identity of the exonuclease-resistant derivative of the sample is known, a finding that the primer has become resistant to exonucleases reveals that the nucleotide(s) present in the polymorphic site of the target molecule is complementary to that of the nucleotide derivative used in the reaction. This method has the advantage that it does not require the determination of large amounts of extraneous sequence data.
  • a solution-based method can be used for determining the identity of a nucleotide of a polymorphic site.
  • Cohen, D. et al. (French Patent 2,650,840; PCT Appln. No. WO91/02087).
  • a primer is employed that is complementary to allelic sequences immediately 3′ to a polymorphic site. The method determines the identity of the nucleotide of that site using labeled dideoxynucleotide derivatives, which, if complementary to the nucleotide of the polymorphic site will become incorporated onto the terminus of the primer.
  • Goelet, P. et al. An alternative method, known as Genetic Bit Analysis or GBA is described by Goelet, P. et al. (PCT Appln. No. 92/15712).
  • the method of Goelet, P. et al. uses mixtures of labeled terminators and a primer that is complementary to the sequence 3′ to a polymorphic site.
  • the labeled terminator that is incorporated is thus determined by, and complementary to, the nucleotide present in the polymorphic site of the target molecule being evaluated.
  • the method of Goelet, P. et al. can be a heterogeneous phase assay, in which the primer or the target molecule is immobilized to a solid phase.
  • oligonucleotides 30-50 bases in length are covalently anchored at the 5′ end to glass cover slips. These anchored strands perform two functions. First, they act as capture sites for the target template strands if the templates are configured with capture tails complementary to the surface-bound oligonucleotides. They also act as primers for the template directed primer extension that forms the basis of the sequence reading.
  • the capture primers function as a fixed position site for sequence determination using multiple cycles of synthesis, detection, and chemical cleavage of the dye-linker to remove the dye. Each cycle consists of adding the polymerase/labeled nucleotide mixture, rinsing, imaging and cleavage of dye.
  • polymerase is modified with a fluorescent donor molecule and immobilized on a glass slide, while each nucleotide is color-coded with an acceptor fluorescent moiety attached to a gamma-phosphate.
  • the system detects the interaction between a fluorescently-tagged polymerase and a fluorescently modified nucleotide as the nucleotide becomes incorporated into the de novo chain.
  • Other sequencing-by-synthesis technologies also exist.
  • any suitable sequencing-by-synthesis platform can be used to identify mutations.
  • four major sequencing-by-synthesis platforms are currently available: the Genome Sequencers from Roche/454 Life Sciences, the 1G Analyzer from Illumina/Solexa, the SOLiD system from Applied BioSystems, and the Heliscope system from Helicos Biosciences. Sequencing-by-synthesis platforms have also been described by Pacific BioSciences and VisiGen Biotechnologies.
  • a plurality of nucleic acid molecules being sequenced is bound to a support (e.g., solid support).
  • a capture sequence/universal priming site can be added at the 3′ and/or 5′ end of the template.
  • the nucleic acids can be bound to the support by hybridizing the capture sequence to a complementary sequence covalently attached to the support.
  • the capture sequence also referred to as a universal capture sequence
  • the capture sequence is a nucleic acid sequence complementary to a sequence attached to a support that may dually serve as a universal primer.
  • a member of a coupling pair (such as, e.g., antibody/antigen, receptor/ligand, or the avidin-biotin pair as described in, e.g., US Patent Application No. 2006/0252077) can be linked to each fragment to be captured on a surface coated with a respective second member of that coupling pair.
  • sequence can be analyzed, for example, by single molecule detection/sequencing, e.g., as described in the Examples and in U.S. Pat. No. 7,283,337, including template-dependent sequencing-by-synthesis.
  • sequencing-by-synthesis the surface-bound molecule is exposed to a plurality of labeled nucleotide triphosphates in the presence of polymerase.
  • the sequence of the template is determined by the order of labeled nucleotides incorporated into the 3′ end of the growing chain. This can be done in real time or can be done in a step-and-repeat mode. For real-time analysis, different optical labels to each nucleotide can be incorporated and multiple lasers can be utilized for stimulation of incorporated nucleotides.
  • Sequencing can also include other massively parallel sequencing or next generation sequencing (NGS) techniques and platforms. Additional examples of massively parallel sequencing techniques and platforms are the Illumina HiSeq or MiSeq, Thermo PGM or Proton, the Pac Bio RS II or Sequel, Qiagen's Gene Reader, and the Oxford Nanopore MinION. Additional similar current massively parallel sequencing technologies can be used, as well as future generations of these technologies.
  • NGS next generation sequencing
  • a DNA or RNA sample can be obtained from a tumor or a bodily fluid, e.g., blood, obtained by known techniques (e.g. venipuncture) or saliva.
  • nucleic acid tests can be performed on dry samples (e.g. hair or skin).
  • a sample can be obtained for sequencing from a tumor and another sample can be obtained from normal tissue for sequencing where the normal tissue is of the same tissue type as the tumor.
  • a sample can be obtained for sequencing from a tumor and another sample can be obtained from normal tissue for sequencing where the normal tissue is of a distinct tissue type relative to the tumor.
  • Tumors can include one or more of lung cancer, melanoma, breast cancer, ovarian cancer, prostate cancer, kidney cancer, gastric cancer, colon cancer, testicular cancer, head and neck cancer, pancreatic cancer, brain cancer, B-cell lymphoma, acute myelogenous leukemia, chronic myelogenous leukemia, chronic lymphocytic leukemia, and T cell lymphocytic leukemia, non-small cell lung cancer, and small cell lung cancer.
  • protein mass spectrometry can be used to identify or validate the presence of mutated peptides bound to MHC proteins on tumor cells.
  • Peptides can be acid-eluted from tumor cells or from HLA molecules that are immunoprecipitated from tumor, and then identified using mass spectrometry.
  • Neoantigens can include nucleotides or polypeptides.
  • a neoantigen can be an RNA sequence that encodes for a polypeptide sequence.
  • Neoantigens useful in vaccines can therefore include nucleotide sequences or polypeptide sequences.
  • Neoantigen peptides can be described in the context of their coding sequence where a neoantigen includes the nucleotide sequence (e.g., DNA or RNA) that codes for the related polypeptide sequence.
  • One or more polypeptides encoded by a neoantigen nucleotide sequence can comprise at least one of: a binding affinity with MHC with an IC50 value of less than 1000 nM, for MHC Class 1 peptides a length of 8-15, 8, 9, 10, 11, 12, 13, 14, or 15 amino acids, presence of sequence motifs within or near the peptide promoting proteasome cleavage, and presence or sequence motifs promoting TAP transport.
  • One or more neoantigens can be presented on the surface of a tumor.
  • One or more neoantigens can be is immunogenic in a subject having a tumor, e.g., capable of eliciting a T cell response or a B cell response in the subject.
  • One or more neoantigens that induce an autoimmune response in a subject can be excluded from consideration in the context of vaccine generation for a subject having a tumor.
  • the size of at least one neoantigenic peptide molecule can comprise, but is not limited to, about 5, about 6, about 7, about 8, about 9, about 10, about 11, about 12, about 13, about 14, about 15, about 16, about 17, about 18, about 19, about 20, about 21, about 22, about 23, about 24, about 25, about 26, about 27, about 28, about 29, about 30, about 31, about 32, about 33, about 34, about 35, about 36, about 37, about 38, about 39, about 40, about 41, about 42, about 43, about 44, about 45, about 46, about 47, about 48, about 49, about 50, about 60, about 70, about 80, about 90, about 100, about 110, about 120 or greater amino molecule residues, and any range derivable therein.
  • the neoantigenic peptide molecules are equal to or less than 50 amino acids.
  • Neoantigenic peptides and polypeptides can be: for MHC Class I 15 residues or less in length and usually consist of between about 8 and about 11 residues, particularly 9 or 10 residues; for MHC Class II, 15-24 residues.
  • a longer peptide can be designed in several ways.
  • a longer peptide could consist of either: (1) individual presented peptides with an extensions of 2-5 amino acids toward the N- and C-terminus of each corresponding gene product; (2) a concatenation of some or all of the presented peptides with extended sequences for each.
  • sequencing reveals a long (>10 residues) neoepitope sequence present in the tumor (e.g.
  • a longer peptide would consist of: (3) the entire stretch of novel tumor-specific amino acids—thus bypassing the need for computational or in vitro test-based selection of the strongest HLA-presented shorter peptide. In both cases, use of a longer peptide allows endogenous processing by patient cells and may lead to more effective antigen presentation and induction of T cell responses.
  • Neoantigenic peptides and polypeptides can be presented on an HLA protein. In some aspects neoantigenic peptides and polypeptides are presented on an HLA protein with greater affinity than a wild-type peptide. In some aspects, a neoantigenic peptide or polypeptide can have an IC50 of at least less than 5000 nM, at least less than 1000 nM, at least less than 500 nM, at least less than 250 nM, at least less than 200 nM, at least less than 150 nM, at least less than 100 nM, at least less than 50 nM or less.
  • neoantigenic peptides and polypeptides do not induce an autoimmune response and/or invoke immunological tolerance when administered to a subject.
  • compositions comprising at least two or more neoantigenic peptides.
  • the composition contains at least two distinct peptides. At least two distinct peptides can be derived from the same polypeptide. By distinct polypeptides is meant that the peptide vary by length, amino acid sequence, or both.
  • the peptides are derived from any polypeptide known to or have been found to contain a tumor specific mutation. Suitable polypeptides from which the neoantigenic peptides can be derived can be found for example in the COSMIC database. COSMIC curates comprehensive information on somatic mutations in human cancer.
  • the peptide contains the tumor specific mutation. In some aspects the tumor specific mutation is a driver mutation for a particular cancer type.
  • Neoantigenic peptides and polypeptides having a desired activity or property can be modified to provide certain desired attributes, e.g., improved pharmacological characteristics, while increasing or at least retaining substantially all of the biological activity of the unmodified peptide to bind the desired MHC molecule and activate the appropriate T cell.
  • desired attributes e.g., improved pharmacological characteristics
  • neoantigenic peptide and polypeptides can be subject to various changes, such as substitutions, either conservative or non-conservative, where such changes might provide for certain advantages in their use, such as improved MHC binding, stability or presentation.
  • conservative substitutions is meant replacing an amino acid residue with another which is biologically and/or chemically similar, e.g., one hydrophobic residue for another, or one polar residue for another.
  • substitutions include combinations such as Gly, Ala; Val, Ile, Leu, Met; Asp, Glu; Asn, Gln; Ser, Thr; Lys, Arg; and Phe, Tyr.
  • the effect of single amino acid substitutions may also be probed using D-amino acids.
  • Such modifications can be made using well known peptide synthesis procedures, as described in e.g., Merrifield, Science 232:341-347 (1986), Barany & Merrifield, The Peptides, Gross & Meienhofer, eds. (N.Y., Academic Press), pp. 1-284 (1979); and Stewart & Young, Solid Phase Peptide Synthesis, (Rockford, Ill., Pierce), 2d Ed. (1984).
  • Modifications of peptides and polypeptides with various amino acid mimetics or unnatural amino acids can be particularly useful in increasing the stability of the peptide and polypeptide in vivo. Stability can be assayed in a number of ways. For instance, peptidases and various biological media, such as human plasma and serum, have been used to test stability. See, e.g., Verhoef et al., Eur. J. Drug Metab Pharmacokin. 11:291-302 (1986). Half-life of the peptides can be conveniently determined using a 25% human serum (v/v) assay. The protocol is generally as follows. Pooled human serum (Type AB, non-heat inactivated) is delipidated by centrifugation before use.
  • Type AB non-heat inactivated
  • the serum is then diluted to 25% with RPMI tissue culture media and used to test peptide stability. At predetermined time intervals a small amount of reaction solution is removed and added to either 6% aqueous trichloracetic acid or ethanol. The cloudy reaction sample is cooled (4 degrees C.) for 15 minutes and then spun to pellet the precipitated serum proteins. The presence of the peptides is then determined by reversed-phase HPLC using stability-specific chromatography conditions.
  • the peptides and polypeptides can be modified to provide desired attributes other than improved serum half-life. For instance, the ability of the peptides to induce CTL activity can be enhanced by linkage to a sequence which contains at least one epitope that is capable of inducing a T helper cell response.
  • Immunogenic peptides/T helper conjugates can be linked by a spacer molecule.
  • the spacer is typically comprised of relatively small, neutral molecules, such as amino acids or amino acid mimetics, which are substantially uncharged under physiological conditions.
  • the spacers are typically selected from, e.g., Ala, Gly, or other neutral spacers of nonpolar amino acids or neutral polar amino acids.
  • the optionally present spacer need not be comprised of the same residues and thus can be a hetero- or homo-oligomer.
  • the spacer will usually be at least one or two residues, more usually three to six residues.
  • the peptide can be linked to the T helper peptide without a spacer.
  • a neoantigenic peptide can be linked to the T helper peptide either directly or via a spacer either at the amino or carboxy terminus of the peptide.
  • the amino terminus of either the neoantigenic peptide or the T helper peptide can be acylated.
  • Exemplary T helper peptides include tetanus toxoid 830-843, influenza 307-319, malaria circumsporozoite 382-398 and 378-389.
  • Proteins or peptides can be made by any technique known to those of skill in the art, including the expression of proteins, polypeptides or peptides through standard molecular biological techniques, the isolation of proteins or peptides from natural sources, or the chemical synthesis of proteins or peptides.
  • the nucleotide and protein, polypeptide and peptide sequences corresponding to various genes have been previously disclosed, and can be found at computerized databases known to those of ordinary skill in the art.
  • One such database is the National Center for Biotechnology Information's Genbank and GenPept databases located at the National Institutes of Health website.
  • the coding regions for known genes can be amplified and/or expressed using the techniques disclosed herein or as would be known to those of ordinary skill in the art.
  • various commercial preparations of proteins, polypeptides and peptides are known to those of skill in the art.
  • a neoantigen includes a nucleic acid (e.g. polynucleotide) that encodes a neoantigenic peptide or portion thereof.
  • the polynucleotide can be, e.g., DNA, cDNA, PNA, CNA, RNA (e.g., mRNA), either single- and/or double-stranded, or native or stabilized forms of polynucleotides, such as, e.g., polynucleotides with a phosphorothiate backbone, or combinations thereof and it may or may not contain introns.
  • a still further aspect provides an expression vector capable of expressing a polypeptide or portion thereof.
  • Expression vectors for different cell types are well known in the art and can be selected without undue experimentation.
  • DNA is inserted into an expression vector, such as a plasmid, in proper orientation and correct reading frame for expression. If necessary, DNA can be linked to the appropriate transcriptional and translational regulatory control nucleotide sequences recognized by the desired host, although such controls are generally available in the expression vector.
  • the vector is then introduced into the host through standard techniques. Guidance can be found e.g. in Sambrook et al. (1989) Molecular Cloning, A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
  • an immunogenic composition e.g., a vaccine composition, capable of raising a specific immune response, e.g., a tumor-specific immune response.
  • Vaccine compositions typically comprise a plurality of neoantigens, e.g., selected using a method described herein. Vaccine compositions can also be referred to as vaccines.
  • a vaccine can contain between 1 and 30 peptides, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, or 30 different peptides, 6, 7, 8, 9, 10 11, 12, 13, or 14 different peptides, or 12, 13 or 14 different peptides.
  • Peptides can include post-translational modifications.
  • a vaccine can contain between 1 and 100 or more nucleotide sequences, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100 or more different nucleotide sequences, 6, 7, 8, 9, 10 11, 12, 13, or 14 different nucleotide sequences, or 12, 13 or 14 different nu
  • a vaccine can contain between 1 and 30 neoantigen sequences, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100 or more different neoantigen sequences, 6, 7, 8, 9, 10 11, 12, 13, or 14 different neoantigen sequences, or 12, 13 or 14 different
  • different peptides and/or polypeptides or nucleotide sequences encoding them are selected so that the peptides and/or polypeptides capable of associating with different MHC molecules, such as different MHC class I molecule.
  • one vaccine composition comprises coding sequence for peptides and/or polypeptides capable of associating with the most frequently occurring MHC class I molecules.
  • vaccine compositions can comprise different fragments capable of associating with at least 2 preferred, at least 3 preferred, or at least 4 preferred MHC class I molecules.
  • the vaccine composition can be capable of raising a specific cytotoxic T-cells response and/or a specific helper T-cell response.
  • a vaccine composition can further comprise an adjuvant and/or a carrier.
  • an adjuvant and/or a carrier examples of useful adjuvants and carriers are given herein below.
  • a composition can be associated with a carrier such as e.g. a protein or an antigen-presenting cell such as e.g. a dendritic cell (DC) capable of presenting the peptide to a T-cell.
  • a carrier such as e.g. a protein or an antigen-presenting cell such as e.g. a dendritic cell (DC) capable of presenting the peptide to a T-cell.
  • DC dendritic cell
  • Adjuvants are any substance whose admixture into a vaccine composition increases or otherwise modifies the immune response to a neoantigen.
  • Carriers can be scaffold structures, for example a polypeptide or a polysaccharide, to which a neoantigen, is capable of being associated.
  • adjuvants are conjugated covalently or non-covalently.
  • an adjuvant to increase an immune response to an antigen is typically manifested by a significant or substantial increase in an immune-mediated reaction, or reduction in disease symptoms.
  • an increase in humoral immunity is typically manifested by a significant increase in the titer of antibodies raised to the antigen
  • an increase in T-cell activity is typically manifested in increased cell proliferation, or cellular cytotoxicity, or cytokine secretion.
  • An adjuvant may also alter an immune response, for example, by changing a primarily humoral or Th response into a primarily cellular, or Th response.
  • Suitable adjuvants include, but are not limited to 1018 ISS, alum, aluminium salts, Amplivax, AS15, BCG, CP-870,893, CpG7909, CyaA, dSLIM, GM-CSF, IC30, IC31, Imiquimod, ImuFact IMP321, IS Patch, ISS, ISCOMATRIX, Juvlmmune, LipoVac, MF59, monophosphoryl lipid A, Montanide IMS 1312, Montanide ISA 206, Montanide ISA 50V, Montanide ISA-51, OK-432, OM-174, OM-197-MP-EC, ONTAK, PepTel vector system, PLG microparticles, resiquimod, SRL172, Virosomes and other Virus-like particles, YF-17D, VEGF trap, R848, beta-glucan, Pam3Cys, Aquila's QS21 stimulon (Aquila Biotech, Worcester
  • Adjuvants such as incomplete Freund's or GM-CSF are useful.
  • GM-CSF Several immunological adjuvants (e.g., MF59) specific for dendritic cells and their preparation have been described previously (Dupuis M, et al., Cell Immunol. 1998; 186(1):18-27; Allison A C; Dev Biol Stand. 1998; 92:3-11).
  • cytokines can be used.
  • cytokines have been directly linked to influencing dendritic cell migration to lymphoid tissues (e.g., TNF-alpha), accelerating the maturation of dendritic cells into efficient antigen-presenting cells for T-lymphocytes (e.g., GM-CSF, IL-1 and IL-4) (U.S. Pat. No. 5,849,589, specifically incorporated herein by reference in its entirety) and acting as immunoadjuvants (e.g., IL-12) (Gabrilovich D I, et al., J Immunother Emphasis Tumor Immunol. 1996 (6):414-418).
  • TLR binding molecules such as RNA binding TLR 7, TLR 8 and/or TLR 9 may also be used.
  • useful adjuvants include, but are not limited to, chemically modified CpGs (e.g. CpR, Idera), Poly(I:C)(e.g. polyi:C12U), non-CpG bacterial DNA or RNA as well as immunoactive small molecules and antibodies such as cyclophosphamide, sunitinib, bevacizumab, celebrex, NCX-4016, sildenafil, tadalafil, vardenafil, sorafinib, XL-999, CP-547632, pazopanib, ZD2171, AZD2171, ipilimumab, tremelimumab, and SC58175, which may act therapeutically and/or as an adjuvant.
  • CpGs e.g. CpR, Idera
  • non-CpG bacterial DNA or RNA as well as immunoactive small molecules and
  • adjuvants and additives can readily be determined by the skilled artisan without undue experimentation.
  • Additional adjuvants include colony-stimulating factors, such as Granulocyte Macrophage Colony Stimulating Factor (GM-CSF, sargramostim).
  • GM-CSF Granulocyte Macrophage Colony Stimulating Factor
  • a vaccine composition can comprise more than one different adjuvant.
  • a therapeutic composition can comprise any adjuvant substance including any of the above or combinations thereof. It is also contemplated that a vaccine and an adjuvant can be administered together or separately in any appropriate sequence.
  • a carrier can be present independently of an adjuvant.
  • the function of a carrier can for example be to increase the molecular weight of in particular mutant to increase activity or immunogenicity, to confer stability, to increase the biological activity, or to increase serum half-life.
  • a carrier can aid presenting peptides to T-cells.
  • a carrier can be any suitable carrier known to the person skilled in the art, for example a protein or an antigen presenting cell.
  • a carrier protein could be but is not limited to keyhole limpet hemocyanin, serum proteins such as transferrin, bovine serum albumin, human serum albumin, thyroglobulin or ovalbumin, immunoglobulins, or hormones, such as insulin or palmitic acid.
  • the carrier is generally a physiologically acceptable carrier acceptable to humans and safe.
  • tetanus toxoid and/or diptheria toxoid are suitable carriers.
  • the carrier can be dextrans for example sepharose.
  • Cytotoxic T-cells recognize an antigen in the form of a peptide bound to an MHC molecule rather than the intact foreign antigen itself.
  • the MHC molecule itself is located at the cell surface of an antigen presenting cell.
  • an activation of CTLs is possible if a trimeric complex of peptide antigen, MHC molecule, and APC is present.
  • it may enhance the immune response if not only the peptide is used for activation of CTLs, but if additionally APCs with the respective MHC molecule are added. Therefore, in some embodiments a vaccine composition additionally contains at least one antigen presenting cell.
  • Neoantigens can also be included in viral vector-based vaccine platforms, such as vaccinia, fowlpox, self-replicating alphavirus, marabavirus, adenovirus (See, e.g., Tatsis et al., Adenoviruses, Molecular Therapy (2004) 10, 616-629), or lentivirus, including but not limited to second, third or hybrid second/third generation lentivirus and recombinant lentivirus of any generation designed to target specific cell types or receptors (See, e.g., Hu et al., Immunization Delivered by Lentiviral Vectors for Cancer and Infectious Diseases, Immunol Rev .
  • this approach can deliver one or more nucleotide sequences that encode one or more neoantigen peptides.
  • the sequences may be flanked by non-mutated sequences, may be separated by linkers or may be preceded with one or more sequences targeting a subcellular compartment (See, e.g., Gros et al., Prospective identification of neoantigen-specific lymphocytes in the peripheral blood of melanoma patients, Nat Med . (2016) 22 (4):433-8, Stronen et al., Targeting of cancer neoantigens with donor-derived T cell receptor repertoires, Science .
  • Truncal peptides meaning those presented by all or most tumor subclones, will be prioritized for inclusion into the vaccine. 53
  • further peptides can be prioritized by estimating the number and identity of tumor subclones and choosing peptides so as to maximize the number of tumor subclones covered by the vaccine. 54
  • an integrated multi-dimensional model can be considered that places candidate neoantigens in a space with at least the following axes and optimizes selection using an integrative approach.
  • a subject has been diagnosed with cancer or is at risk of developing cancer.
  • a subject can be a human, dog, cat, horse or any animal in which a tumor specific immune response is desired.
  • a tumor can be any solid tumor such as breast, ovarian, prostate, lung, kidney, gastric, colon, testicular, head and neck, pancreas, brain, melanoma, and other tumors of tissue organs and hematological tumors, such as lymphomas and leukemias, including acute myelogenous leukemia, chronic myelogenous leukemia, chronic lymphocytic leukemia, T cell lymphocytic leukemia, and B cell lymphomas.
  • a neoantigen can be administered in an amount sufficient to induce a CTL response.
  • a neoantigen can be administered alone or in combination with other therapeutic agents.
  • the therapeutic agent is for example, a chemotherapeutic agent, radiation, or immunotherapy. Any suitable therapeutic treatment for a particular cancer can be administered.
  • a subject can be further administered an anti-immunosuppressive/immunostimulatory agent such as a checkpoint inhibitor.
  • an anti-immunosuppressive/immunostimulatory agent such as a checkpoint inhibitor.
  • the subject can be further administered an anti-CTLA antibody or anti-PD-1 or anti-PD-L1.
  • Blockade of CTLA-4 or PD-L1 by antibodies can enhance the immune response to cancerous cells in the patient.
  • CTLA-4 blockade has been shown effective when following a vaccination protocol.
  • a neoantigen or its variant can be prepared for intravenous (i.v.) injection, sub-cutaneous (s.c.) injection, intradermal (i.d.) injection, intraperitoneal (i.p.) injection, intramuscular (i.m.) injection.
  • Methods of injection include s.c., i.d., i.p., i.m., and i.v.
  • Methods of DNA or RNA injection include i.d., i.m., s.c., i.p. and i.v.
  • Other methods of administration of the vaccine composition are known to those skilled in the art.
  • a vaccine can be compiled so that the selection, number and/or amount of neoantigens present in the composition is/are tissue, cancer, and/or patient-specific. For instance, the exact selection of peptides can be guided by expression patterns of the parent proteins in a given tissue. The selection can be dependent on the specific type of cancer, the status of the disease, earlier treatment regimens, the immune status of the patient, and, of course, the HLA-haplotype of the patient. Furthermore, a vaccine can contain individualized components, according to personal needs of the particular patient. Examples include varying the selection of neoantigens according to the expression of the neoantigen in the particular patient or adjustments for secondary treatments following a first round or scheme of treatment.
  • neoantigens with similar normal self-peptides that are expressed in high amounts in normal tissues can be avoided or be present in low amounts in a composition described herein.
  • the respective pharmaceutical composition for treatment of this cancer can be present in high amounts and/or more than one neoantigen specific for this particularly neoantigen or pathway of this neoantigen can be included.
  • compositions comprising a neoantigen can be administered to an individual already suffering from cancer.
  • compositions are administered to a patient in an amount sufficient to elicit an effective CTL response to the tumor antigen and to cure or at least partially arrest symptoms and/or complications.
  • An amount adequate to accomplish this is defined as “therapeutically effective dose.” Amounts effective for this use will depend on, e.g., the composition, the manner of administration, the stage and severity of the disease being treated, the weight and general state of health of the patient, and the judgment of the prescribing physician. It should be kept in mind that compositions can generally be employed in serious disease states, that is, life-threatening or potentially life threatening situations, especially when the cancer has metastasized. In such cases, in view of the minimization of extraneous substances and the relative nontoxic nature of a neoantigen, it is possible and can be felt desirable by the treating physician to administer substantial excesses of these compositions.
  • administration can begin at the detection or surgical removal of tumors. This is followed by boosting doses until at least symptoms are substantially abated and for a period thereafter.
  • compositions for therapeutic treatment are intended for parenteral, topical, nasal, oral or local administration.
  • a pharmaceutical compositions can be administered parenterally, e.g., intravenously, subcutaneously, intradermally, or intramuscularly.
  • the compositions can be administered at the site of surgical excision to induce a local immune response to the tumor.
  • compositions for parenteral administration which comprise a solution of the neoantigen and vaccine compositions are dissolved or suspended in an acceptable carrier, e.g., an aqueous carrier.
  • aqueous carriers can be used, e.g., water, buffered water, 0.9% saline, 0.3% glycine, hyaluronic acid and the like. These compositions can be sterilized by conventional, well known sterilization techniques, or can be sterile filtered. The resulting aqueous solutions can be packaged for use as is, or lyophilized, the lyophilized preparation being combined with a sterile solution prior to administration.
  • compositions may contain pharmaceutically acceptable auxiliary substances as required to approximate physiological conditions, such as pH adjusting and buffering agents, tonicity adjusting agents, wetting agents and the like, for example, sodium acetate, sodium lactate, sodium chloride, potassium chloride, calcium chloride, sorbitan monolaurate, triethanolamine oleate, etc.
  • auxiliary substances such as pH adjusting and buffering agents, tonicity adjusting agents, wetting agents and the like, for example, sodium acetate, sodium lactate, sodium chloride, potassium chloride, calcium chloride, sorbitan monolaurate, triethanolamine oleate, etc.
  • Neoantigens can also be administered via liposomes, which target them to a particular cells tissue, such as lymphoid tissue. Liposomes are also useful in increasing half-life. Liposomes include emulsions, foams, micelles, insoluble monolayers, liquid crystals, phospholipid dispersions, lamellar layers and the like. In these preparations the neoantigen to be delivered is incorporated as part of a liposome, alone or in conjunction with a molecule which binds to, e.g., a receptor prevalent among lymphoid cells, such as monoclonal antibodies which bind to the CD45 antigen, or with other therapeutic or immunogenic compositions.
  • a receptor prevalent among lymphoid cells such as monoclonal antibodies which bind to the CD45 antigen, or with other therapeutic or immunogenic compositions.
  • liposomes filled with a desired neoantigen can be directed to the site of lymphoid cells, where the liposomes then deliver the selected therapeutic/immunogenic compositions.
  • Liposomes can be formed from standard vesicle-forming lipids, which generally include neutral and negatively charged phospholipids and a sterol, such as cholesterol. The selection of lipids is generally guided by consideration of, e.g., liposome size, acid lability and stability of the liposomes in the blood stream. A variety of methods are available for preparing liposomes, as described in, e.g., Szoka et al., Ann. Rev. Biophys. Bioeng. 9; 467 (1980), U.S. Pat. Nos. 4,235,871, 4,501,728, 4,501,728, 4,837,028, and 5,019,369.
  • a ligand to be incorporated into the liposome can include, e.g., antibodies or fragments thereof specific for cell surface determinants of the desired immune system cells.
  • a liposome suspension can be administered intravenously, locally, topically, etc. in a dose which varies according to, inter alia, the manner of administration, the peptide being delivered, and the stage of the disease being treated.
  • nucleic acids encoding a peptide and optionally one or more of the peptides described herein can also be administered to the patient.
  • a number of methods are conveniently used to deliver the nucleic acids to the patient.
  • the nucleic acid can be delivered directly, as “naked DNA”. This approach is described, for instance, in Wolff et al., Science 247: 1465-1468 (1990) as well as U.S. Pat. Nos. 5,580,859 and 5,589,466.
  • the nucleic acids can also be administered using ballistic delivery as described, for instance, in U.S. Pat. No. 5,204,253.
  • Particles comprised solely of DNA can be administered.
  • DNA can be adhered to particles, such as gold particles.
  • Approaches for delivering nucleic acid sequences can include viral vectors, mRNA vectors, and DNA vectors with or without electroporation.
  • the nucleic acids can also be delivered complexed to cationic compounds, such as cationic lipids.
  • cationic compounds such as cationic lipids.
  • Lipid-mediated gene delivery methods are described, for instance, in 9618372WOAWO 96/18372; 9324640WOAWO 93/24640; Mannino & Gould-Fogerite, BioTechniques 6(7): 682-691 (1988); U.S. Pat. No. 5,279,833 Rose U.S. Pat. Nos. 5,279,833; 9,106,309WOAWO 91/06309; and Felgner et al., Proc. Natl. Acad. Sci. USA 84: 7413-7414 (1987).
  • Neoantigens can also be included in viral vector-based vaccine platforms, such as vaccinia, fowlpox, self-replicating alphavirus, marabavirus, adenovirus (See, e.g., Tatsis et al., Adenoviruses, Molecular Therapy (2004) 10, 616-629), or lentivirus, including but not limited to second, third or hybrid second/third generation lentivirus and recombinant lentivirus of any generation designed to target specific cell types or receptors (See, e.g., Hu et al., Immunization Delivered by Lentiviral Vectors for Cancer and Infectious Diseases, Immunol Rev .
  • this approach can deliver one or more nucleotide sequences that encode one or more neoantigen peptides.
  • the sequences may be flanked by non-mutated sequences, may be separated by linkers or may be preceded with one or more sequences targeting a subcellular compartment (See, e.g., Gros et al., Prospective identification of neoantigen-specific lymphocytes in the peripheral blood of melanoma patients, Nat Med . (2016) 22 (4):433-8, Stronen et al., Targeting of cancer neoantigens with donor-derived T cell receptor repertoires, Science .
  • a means of administering nucleic acids uses minigene constructs encoding one or multiple epitopes.
  • These epitope-encoding DNA sequences are directly adjoined, creating a continuous polypeptide sequence.
  • additional elements can be incorporated into the minigene design. Examples of amino acid sequence that could be reverse translated and included in the minigene sequence include: helper T lymphocyte, epitopes, a leader (signal) sequence, and an endoplasmic reticulum retention signal.
  • MHC presentation of CTL epitopes can be improved by including synthetic (e.g. poly-alanine) or naturally-occurring flanking sequences adjacent to the CTL epitopes.
  • the minigene sequence is converted to DNA by assembling oligonucleotides that encode the plus and minus strands of the minigene. Overlapping oligonucleotides (30-100 bases long) are synthesized, phosphorylated, purified and annealed under appropriate conditions using well known techniques. The ends of the oligonucleotides are joined using T4 DNA ligase. This synthetic minigene, encoding the CTL epitope polypeptide, can then cloned into a desired expression vector.
  • Purified plasmid DNA can be prepared for injection using a variety of formulations. The simplest of these is reconstitution of lyophilized DNA in sterile phosphate-buffer saline (PBS). A variety of methods have been described, and new techniques can become available. As noted above, nucleic acids are conveniently formulated with cationic lipids. In addition, glycolipids, fusogenic liposomes, peptides and compounds referred to collectively as protective, interactive, non-condensing (PINC) could also be complexed to purified plasmid DNA to influence variables such as stability, intramuscular dispersion, or trafficking to specific organs or cell types.
  • PINC protective, interactive, non-condensing
  • Also disclosed is a method of manufacturing a tumor vaccine comprising performing the steps of a method disclosed herein; and producing a tumor vaccine comprising a plurality of neoantigens or a subset of the plurality of neoantigens.
  • Neoantigens disclosed herein can be manufactured using methods known in the art.
  • a method of producing a neoantigen or a vector (e.g., a vector including at least one sequence encoding one or more neoantigens) disclosed herein can include culturing a host cell under conditions suitable for expressing the neoantigen or vector wherein the host cell comprises at least one polynucleotide encoding the neoantigen or vector, and purifying the neoantigen or vector.
  • Standard purification methods include chromatographic techniques, electrophoretic, immunological, precipitation, dialysis, filtration, concentration, and chromatofocusing techniques.
  • Host cells can include a Chinese Hamster Ovary (CHO) cell, NS0 cell, yeast, or a HEK293 cell.
  • Host cells can be transformed with one or more polynucleotides comprising at least one nucleic acid sequence that encodes a neoantigen or vector disclosed herein, optionally wherein the isolated polynucleotide further comprises a promoter sequence operably linked to the at least one nucleic acid sequence that encodes the neoantigen or vector.
  • the isolated polynucleotide can be cDNA.
  • Improvements in analysis methods address the suboptimal sensitivity and specificity of common research mutation calling approaches, and specifically consider customizations relevant for neoantigen identification in the clinical setting. These include:
  • RNA CoMPASS 44 In samples with poly-adenylated RNA, the presence of viral and microbial RNA in the RNA-seq data will be assessed using RNA CoMPASS 44 or a similar method, toward the identification of additional factors that may predict patient response.
  • IP immunoprecipitation
  • Immunoprecipitation was performed using antibodies coupled to beads where the antibody is specific for HLA molecules.
  • a pan-Class I HLA immunoprecipitation a pan-Class I CR antibody is used, for Class II HLA-DR, an HLA-DR antibody is used.
  • Antibody is covalently attached to NHS-sepharose beads during overnight incubation. After covalent attachment, the beads were washed and aliquoted for IP. (59, 60)
  • the clarified tissue lysate is added to the antibody beads for the immunoprecipitation.
  • the beads are removed from the lysate and the lysate stored for additional experiments, including additional IPs.
  • the IP beads are washed to remove non-specific binding and the HLA/peptide complex is eluted from the beads using standard techniques.
  • the protein components are removed from the peptides using a molecular weight spin column or C18 fractionation.
  • the resultant peptides are taken to dryness by SpeedVac evaporation and in some instances are stored at ⁇ 20 C prior to MS analysis.
  • Dried peptides are reconstituted in an HPLC buffer suitable for reverse phase chromatography and loaded onto a C-18 microcapillary HPLC column for gradient elution in a Fusion Lumos mass spectrometer (Thermo).
  • MS1 spectra of peptide mass/charge (m/z) were collected in the Orbitrap detector at high resolution followed by MS2 low resolution scans collected in the ion trap detector after HCD fragmentation of the selected ion.
  • MS2 spectra can be obtained using either CID or ETD fragmentation methods or any combination of the three techniques to attain greater amino acid coverage of the peptide.
  • MS2 spectra can also be measured with high resolution mass accuracy in the Orbitrap detector.
  • MS2 spectra from each analysis are searched against a protein database using Comet (61, 62) and the peptide identification are scored using Percolator (63-65).
  • FIG. 2A is an overview of an environment 100 for identifying likelihoods of peptide presentation in patients, in accordance with an embodiment.
  • the environment 100 provides context in order to introduce a presentation identification system 160 , itself including a presentation information store 165 .
  • the presentation identification system 160 is one or computer models, embodied in a computing system as discussed below with respect to FIG. 14 , that receives peptide sequences associated with a set of MHC alleles and determines likelihoods that the peptide sequences will be presented by one or more of the set of associated MHC alleles. This is useful in a variety of contexts.
  • One specific use case for the presentation identification system 160 is that it is able to receive nucleotide sequences of candidate neoantigens associated with a set of MHC alleles from tumor cells of a patient 110 and determine likelihoods that the candidate neoantigens will be presented by one or more of the associated MHC alleles of the tumor and/or induce immunogenic responses in the immune system of the patient 110 .
  • Those candidate neoantigens with high likelihoods as determined by system 160 can be selected for inclusion in a vaccine 118 , such an anti-tumor immune response can be elicited from the immune system of the patient 110 providing the
  • the presentation identification system 160 determines presentation likelihoods through one or more presentation models. Specifically, the presentation models generate likelihoods of whether given peptide sequences will be presented for a set of associated MHC alleles, and are generated based on presentation information stored in store 165 . For example, the presentation models may generate likelihoods of whether a peptide sequence “YVYVADVAAK (SEQ ID NO: 1)” will be presented for the set of alleles HLA-A*02:01, HLA-B*07:02, HLA-B*08:03, HLA-C*01:04, HLA-A*06:03, HLA-B*01:04 on the cell surface of the sample.
  • YVYVADVAAK SEQ ID NO: 1
  • the presentation information 165 contains information on whether peptides bind to different types of MHC alleles such that those peptides are presented by MHC alleles, which in the models is determined depending on positions of amino acids in the peptide sequences.
  • the presentation model can predict whether an unrecognized peptide sequence will be presented in association with an associated set of MHC alleles based on the presentation information 165 .
  • FIG. 2 illustrates a method of obtaining presentation information, in accordance with an embodiment.
  • the presentation information 165 includes two general categories of information: allele-interacting information and allele-noninteracting information. Allele-interacting information includes information that influence presentation of peptide sequences that are dependent on the type of MHC allele. Allele-noninteracting information includes information that influence presentation of peptide sequences that are independent on the type of MHC allele.
  • Allele-interacting information primarily includes identified peptide sequences that are known to have been presented by one or more identified MHC molecules from humans, mice, etc. Notably, this may or may not include data obtained from tumor samples.
  • the presented peptide sequences may be identified from cells that express a single MHC allele. In this case the presented peptide sequences are generally collected from single-allele cell lines that are engineered to express a predetermined MHC allele and that are subsequently exposed to synthetic protein. Peptides presented on the MHC allele are isolated by techniques such as acid-elution and identified through mass spectrometry. FIG.
  • FIG. 2B shows an example of this, where the example peptide YEMFNDKS (SEQ ID NO: 2), presented on the predetermined MHC allele HLA-A*01:01, is isolated and identified through mass spectrometry. Since in this situation peptides are identified through cells engineered to express a single predetermined MHC protein, the direct association between a presented peptide and the MHC protein to which it was bound to is definitively known.
  • SEQ ID NO: 2B shows an example of this, where the example peptide YEMFNDKS (SEQ ID NO: 2), presented on the predetermined MHC allele HLA-A*01:01, is isolated and identified through mass spectrometry. Since in this situation peptides are identified through cells engineered to express a single predetermined MHC protein, the direct association between a presented peptide and the MHC protein to which it was bound to is definitively known.
  • the presented peptide sequences may also be collected from cells that express multiple MHC alleles. Typically in humans, 6 different types of MHC molecules are expressed for a cell. Such presented peptide sequences may be identified from multiple-allele cell lines that are engineered to express multiple predetermined MHC alleles. Such presented peptide sequences may also be identified from tissue samples, either from normal tissue samples or tumor tissue samples. In this case particularly, the MHC molecules can be immunoprecipitated from normal or tumor tissue. Peptides presented on the multiple MHC alleles can similarly be isolated by techniques such as acid-elution and identified through mass spectrometry. FIG.
  • 2C shows an example of this, where the six example peptides, YEMFNDKSF (SEQ ID NO: 3), HROEIFSHDFJ (SEQ ID NO: 4), FJIEJFOESS (SEQ ID NO: 5), NEIOREIREI (SEQ ID NO: 6), JFKSIFEMMSJDSSU (SEQ ID NO: 7), and KNFLENFIESOFI (SEQ ID NO: 8), are presented on identified MHC alleles HLA-A*01:01, HLA-A*02:01, HLA-B*07:02, HLA-B*08:01, HLA-C*01:03, and HLA-C*01:04 and are isolated and identified through mass spectrometry.
  • the direct association between a presented peptide and the MHC protein to which it was bound to may be unknown since the bound peptides are isolated from the MHC molecules before being identified.
  • Allele-interacting information can also include mass spectrometry ion current which depends on both the concentration of peptide-MHC molecule complexes, and the ionization efficiency of peptides.
  • the ionization efficiency varies from peptide to peptide in a sequence-dependent manner. Generally, ionization efficiency varies from peptide to peptide over approximately two orders of magnitude, while the concentration of peptide-MHC complexes varies over a larger range than that.
  • Allele-interacting information can also include measurements or predictions of binding affinity between a given MHC allele and a given peptide.
  • One or more affinity models can generate such predictions.
  • presentation information 165 may include a binding affinity prediction of 1000 nM between the peptide YEMFNDKSF (SEQ ID NO: 3) and the allele HLA-A*01:01. Few peptides with IC50>1000 nm are presented by the MHC, and lower IC50 values increase the probability of presentation.
  • Allele-interacting information can also include measurements or predictions of stability of the MHC complex.
  • One or more stability models that can generate such predictions. More stable peptide-MHC complexes (i.e., complexes with longer half-lives) are more likely to be presented at high copy number on tumor cells and on antigen-presenting cells that encounter vaccine antigen.
  • presentation information 165 may include a stability prediction of a half-life of 1 h for the molecule HLA-A*01:01.
  • Allele-interacting information can also include the measured or predicted rate of the formation reaction for the peptide-MHC complex. Complexes that form at a higher rate are more likely to be presented on the cell surface at high concentration.
  • Allele-interacting information can also include the sequence and length of the peptide.
  • MHC class I molecules typically prefer to present peptides with lengths between 8 and 15 peptides. 60-80% of presented peptides have length 9 . Histograms of presented peptide lengths from several cell lines are shown in FIG. 5 .
  • Allele-interacting information can also include the presence of kinase sequence motifs on the neoantigen encoded peptide, and the absence or presence of specific post-translational modifications on the neoantigen encoded peptide.
  • the presence of kinase motifs affects the probability of post-translational modification, which may enhance or interfere with MHC binding.
  • Allele-interacting information can also include the expression or activity levels of proteins involved in the process of post-translational modification, e.g., kinases (as measured or predicted from RNA seq, mass spectrometry, or other methods).
  • proteins involved in the process of post-translational modification e.g., kinases (as measured or predicted from RNA seq, mass spectrometry, or other methods).
  • Allele-interacting information can also include the probability of presentation of peptides with similar sequence in cells from other individuals expressing the particular MHC allele as assessed by mass-spectrometry proteomics or other means.
  • Allele-interacting information can also include the expression levels of the particular MHC allele in the individual in question (e.g. as measured by RNA-seq or mass spectrometry). Peptides that bind most strongly to an MHC allele that is expressed at high levels are more likely to be presented than peptides that bind most strongly to an MHC allele that is expressed at a low level.
  • Allele-interacting information can also include the overall neoantigen encoded peptide-sequence-independent probability of presentation by the particular MHC allele in other individuals who express the particular MHC allele.
  • Allele-interacting information can also include the overall peptide-sequence-independent probability of presentation by MHC alleles in the same family of molecules (e.g., HLA-A, HLA-B, HLA-C, HLA-DQ, HLA-DR, HLA-DP) in other individuals.
  • HLA-C molecules are typically expressed at lower levels than HLA-A or HLA-B molecules, and consequently, presentation of a peptide by HLA-C is a priori less probable than presentation by HLA-A or HLA-B 11.
  • Allele-interacting information can also include the protein sequence of the particular MHC allele.
  • Any MHC allele-noninteracting information listed in the below section can also be modeled as an MHC allele-interacting information.
  • Allele-noninteracting information can include C-terminal sequences flanking the neoantigen encoded peptide within its source protein sequence.
  • C-terminal flanking sequences may impact proteasomal processing of peptides.
  • the C-terminal flanking sequence is cleaved from the peptide by the proteasome before the peptide is transported to the endoplasmic reticulum and encounters MHC alleles on the surfaces of cells. Consequently, MHC molecules receive no information about the C-terminal flanking sequence, and thus, the effect of the C-terminal flanking sequence cannot vary depending on MHC allele type. For example, going back to the example shown in FIG.
  • presentation information 165 may include the C-terminal flanking sequence FOEIFNDKSLDKFJI (SEQ ID NO: 9) of the presented peptide FJIEJFOESS (SEQ ID NO: 5) identified from the source protein of the peptide.
  • Allele-noninteracting information can also include mRNA quantification measurements.
  • mRNA quantification data can be obtained for the same samples that provide the mass spectrometry training data. As later described in reference to FIG. 13H , RNA expression was identified to be a strong predictor of peptide presentation.
  • the mRNA quantification measurements are identified from software tool RSEM. Detailed implementation of the RSEM software tool can be found at Bo Li and Colin N. Dewey. RSEM: accurate transcript quantification from RNA - Seq data with or without a reference genome . BMC Bioinformatics, 12:323, August 2011. In one embodiment, the mRNA quantification is measured in units of fragments per kilobase of transcript per Million mapped reads (FPKM).
  • FPKM Million mapped reads
  • Allele-noninteracting information can also include the N-terminal sequences flanking the peptide within its source protein sequence.
  • Allele-noninteracting information can also include the presence of protease cleavage motifs in the peptide, optionally weighted according to the expression of corresponding proteases in the tumor cells (as measured by RNA-seq or mass spectrometry). Peptides that contain protease cleavage motifs are less likely to be presented, because they will be more readily degraded by proteases, and will therefore be less stable within the cell.
  • Allele-noninteracting information can also include the turnover rate of the source protein as measured in the appropriate cell type. Faster turnover rate (i.e., lower half-life) increases the probability of presentation; however, the predictive power of this feature is low if measured in a dissimilar cell type.
  • Allele-noninteracting information can also include the length of the source protein, optionally considering the specific splice variants (“isoforms”) most highly expressed in the tumor cells as measured by RNA-seq or proteome mass spectrometry, or as predicted from the annotation of germline or somatic splicing mutations detected in DNA or RNA sequence data.
  • Allele-noninteracting information can also include the level of expression of the proteasome, immunoproteasome, thymoproteasome, or other proteases in the tumor cells (which may be measured by RNA-seq, proteome mass spectrometry, or immunohistochemistry). Different proteasomes have different cleavage site preferences. More weight will be given to the cleavage preferences of each type of proteasome in proportion to its expression level.
  • Allele-noninteracting information can also include the expression of the source gene of the peptide (e.g., as measured by RNA-seq or mass spectrometry). Possible optimizations include adjusting the measured expression to account for the presence of stromal cells and tumor-infiltrating lymphocytes within the tumor sample. Peptides from more highly expressed genes are more likely to be presented. Peptides from genes with undetectable levels of expression can be excluded from consideration.
  • Allele-noninteracting information can also include the probability that the source mRNA of the neoantigen encoded peptide will be subject to nonsense-mediated decay as predicted by a model of nonsense-mediated decay, for example, the model from Rivas et al, Science 2015.
  • Allele-noninteracting information can also include the typical tissue-specific expression of the source gene of the peptide during various stages of the cell cycle. Genes that are expressed at a low level overall (as measured by RNA-seq or mass spectrometry proteomics) but that are known to be expressed at a high level during specific stages of the cell cycle are likely to produce more presented peptides than genes that are stably expressed at very low levels.
  • Allele-noninteracting information can also include a comprehensive catalog of features of the source protein as given in e.g. uniProt or PDB http://www.rcsb.org/pdb/home/home.do. These features may include, among others: the secondary and tertiary structures of the protein, subcellular localization 11, Gene ontology (GO) terms. Specifically, this information may contain annotations that act at the level of the protein, e.g., 5′ UTR length, and annotations that act at the level of specific residues, e.g., helix motif between residues 300 and 310. These features can also include turn motifs, sheet motifs, and disordered residues.
  • Allele-noninteracting information can also include features describing the properties of the domain of the source protein containing the peptide, for example: secondary or tertiary structure (e.g., alpha helix vs beta sheet); Alternative splicing.
  • Allele-noninteracting information can also include features describing the presence or absence of a presentation hotspot at the position of the peptide in the source protein of the peptide.
  • Allele-noninteracting information can also include the probability of presentation of peptides from the source protein of the peptide in question in other individuals (after adjusting for the expression level of the source protein in those individuals and the influence of the different HLA types of those individuals).
  • Allele-noninteracting information can also include the probability that the peptide will not be detected or over-represented by mass spectrometry due to technical biases.
  • a gene expression assay such as RNASeq, microarray(s), targeted panel(s) such as Nanostring, or single/multi-gene representatives of gene modules measured by assays such as RT-PCR (which need not contain the source protein of the peptide) that are informative about the state of the tumor cells, stroma, or tumor-infiltrating lymphocytes (TILs).
  • Allele-noninteracting information can also include the copy number of the source gene of the peptide in the tumor cells.
  • peptides from genes that are subject to homozygous deletion in tumor cells can be assigned a probability of presentation of zero.
  • Allele-noninteracting information can also include the probability that the peptide binds to the TAP or the measured or predicted binding affinity of the peptide to the TAP. Peptides that are more likely to bind to the TAP, or peptides that bind the TAP with higher affinity are more likely to be presented.
  • Allele-noninteracting information can also include the expression level of TAP in the tumor cells (which may be measured by RNA-seq, proteome mass spectrometry, immunohistochemistry). Higher TAP expression levels increase the probability of presentation of all peptides.
  • Allele-noninteracting information can also include the presence or absence of tumor mutations, including, but not limited to:
  • germline polymorphisms including, but not limited to:
  • Allele-noninteracting information can also include tumor type (e.g., NSCLC, melanoma).
  • tumor type e.g., NSCLC, melanoma
  • Allele-noninteracting information can also include known functionality of HLA alleles, as reflected by, for instance HLA allele suffixes.
  • HLA allele suffixes For example, the N suffix in the allele name HLA-A*24:09N indicates a null allele that is not expressed and is therefore unlikely to present epitopes; the full HLA allele suffix nomenclature is described at https://www.ebi.ac.uk/ipd/imgt/hla/nomenclature/suffixes.html.
  • Allele-noninteracting information can also include clinical tumor subtype (e.g., squamous lung cancer vs. non-squamous).
  • clinical tumor subtype e.g., squamous lung cancer vs. non-squamous.
  • Allele-noninteracting information can also include smoking history.
  • Allele-noninteracting information can also include history of sunburn, sun exposure, or exposure to other mutagens.
  • Allele-noninteracting information can also include the typical expression of the source gene of the peptide in the relevant tumor type or clinical subtype, optionally stratified by driver mutation. Genes that are typically expressed at high levels in the relevant tumor type are more likely to be presented.
  • Allele-noninteracting information can also include the frequency of the mutation in all tumors, or in tumors of the same type, or in tumors from individuals with at least one shared MHC allele, or in tumors of the same type in individuals with at least one shared MHC allele.
  • the list of features used to predict a probability of presentation may also include the annotation of the mutation (e.g., missense, read-through, frameshift, fusion, etc.) or whether the mutation is predicted to result in nonsense-mediated decay (NMD).
  • NMD nonsense-mediated decay
  • peptides from protein segments that are not translated in tumor cells due to homozygous early-stop mutations can be assigned a probability of presentation of zero. NMD results in decreased mRNA translation, which decreases the probability of presentation.
  • FIG. 3 is a high-level block diagram illustrating the computer logic components of the presentation identification system 160 , according to one embodiment.
  • the presentation identification system 160 includes a data management module 312 , an encoding module 314 , a training module 316 , and a prediction module 320 .
  • the presentation identification system 160 is also comprised of a training data store 170 and a presentation models store 175 .
  • Some embodiments of the model management system 160 have different modules than those described here. Similarly, the functions can be distributed among the modules in a different manner than is described here.
  • the data management module 312 generates sets of training data 170 from the presentation information 165 .
  • Each set of training data contains a plurality of data instances, in which each data instance i contains a set of independent variables z i that include at least a presented or non-presented peptide sequence p i , one or more associated MHC alleles a i associated with the peptide sequence p i , and a dependent variable y i that represents information that the presentation identification system 160 is interested in predicting for new values of independent variables.
  • the dependent variable y i is a binary label indicating whether peptide p i was presented by the one or more associated MHC alleles a i .
  • the dependent variable y i can represent any other kind of information that the presentation identification system 160 is interested in predicting dependent on the independent variables z i .
  • the dependent variable y i may also be a numerical value indicating the mass spectrometry ion current identified for the data instance.
  • the peptide sequence p i for data instance i is a sequence of k i amino acids, in which k i may vary between data instances i within a range. For example, that range may be 8-15 for MHC class I or 9-30 for MHC class II.
  • all peptide sequences p i in a training data set may have the same length, e.g. 9.
  • the number of amino acids in a peptide sequence may vary depending on the type of MHC alleles (e.g., MHC alleles in humans, etc.).
  • the MHC alleles a i for data instance i indicate which MHC alleles were present in association with the corresponding peptide sequence p i .
  • the data management module 312 may also include additional allele-interacting variables, such as binding affinity b i and stability s i predictions in conjunction with the peptide sequences p i and associated MHC alleles a i contained in the training data 170 .
  • the training data 170 may contain binding affinity predictions b i between a peptide p i and each of the associated MHC molecules indicated in a i .
  • the training data 170 may contain stability predictions s i for each of the MHC alleles indicated in a i .
  • the data management module 312 may also include allele-noninteracting variables w i , such as C-terminal flanking sequences and mRNA quantification measurements in conjunction with the peptide sequences p i .
  • the data management module 312 also identifies peptide sequences that are not presented by MHC alleles to generate the training data 170 . Generally, this involves identifying the “longer” sequences of source protein that include presented peptide sequences prior to presentation. When the presentation information contains engineered cell lines, the data management module 312 identifies a series of peptide sequences in the synthetic protein to which the cells were exposed to that were not presented on MHC alleles of the cells. When the presentation information contains tissue samples, the data management module 312 identifies source proteins from which presented peptide sequences originated from, and identifies a series of peptide sequences in the source protein that were not presented on MHC alleles of the tissue sample cells.
  • the data management module 312 may also artificially generate peptides with random sequences of amino acids and identify the generated sequences as peptides not presented on MHC alleles. This can be accomplished by randomly generating peptide sequences allows the data management module 312 to easily generate large amounts of synthetic data for peptides not presented on MHC alleles. Since in reality, a small percentage of peptide sequences are presented by MHC alleles, the synthetically generated peptide sequences are highly likely not to have been presented by MHC alleles even if they were included in proteins processed by cells.
  • FIG. 4 illustrates an example set of training data 170 A, according to one embodiment.
  • the first 3 data instances in the training data 170 A indicate peptide presentation information from a single-allele cell line involving the allele HLA-C*01:03 and 3 peptide sequences QCEIOWARE (SEQ ID NO: 10), FIEUHFWI (SEQ ID NO: 11), and FEWRHRJTRUJR (SEQ ID NO: 12).
  • the fourth data instance in the training data 170 A indicates peptide information from a multiple-allele cell line involving the alleles HLA-B*07:02, HLA-C*01:03, HLA-A*01:01 and a peptide sequence QIEJOEIJE (SEQ ID NO: 13).
  • the first data instance indicates that peptide sequence QCEIOWARE (SEQ ID NO: 10) was not presented by the allele HLA-C*01:03.
  • the peptide sequence may be randomly generated by the data management module 312 or identified from source protein of presented peptides.
  • the training data 170 A also includes a binding affinity prediction of 1000 nM and a stability prediction of a half-life of 1 h for the peptide sequence-allele pair.
  • the training data 170 A also includes allele-noninteracting variables, such as the C-terminal flanking sequence of the peptide FJELFISBOSJFIE (SEQ ID NO: 14), and a mRNA quantification measurement of 10 2 FPKM.
  • the fourth data instance indicates that peptide sequence QIEJOEIJE (SEQ ID NO: 13) was presented by one of the alleles HLA-B*07:02, HLA-C*01:03, or HLA-A*01:01.
  • the training data 170 A also includes binding affinity predictions and stability predictions for each of the alleles, as well as the C-flanking sequence of the peptide and the mRNA quantification measurement for the peptide.
  • the encoding module 314 encodes information contained in the training data 170 into a numerical representation that can be used to generate the one or more presentation models.
  • the encoding module 314 one-hot encodes sequences (e.g., peptide sequences or C-terminal flanking sequences) over a predetermined 20-letter amino acid alphabet.
  • sequences e.g., peptide sequences or C-terminal flanking sequences
  • a peptide sequence p i with k i amino acids is represented as a row vector of 20 ⁇ k i elements, where a single element among p i 20 ⁇ (j ⁇ 1)+1 , p i 20 ⁇ (j ⁇ 1)+2 , . . .
  • the encoding module 314 may further encode the peptides into equal-length vectors by adding a PAD character to extend the predetermined alphabet. For example, this may be performed by left-padding the peptide sequences with the PAD character until the length of the peptide sequence reaches the peptide sequence with the greatest length in the training data 170 .
  • the encoding module 314 numerically represents each sequence as a row vector of (20+1) ⁇ k max elements.
  • each independent variable or column in the peptide sequence p i or c i represents presence of a particular amino acid at a particular position of the sequence.
  • sequence data was described in reference to sequences having amino acid sequences, the method can similarly be extended to other types of sequence data, such as DNA or RNA sequence data, and the like.
  • the elements corresponding to the MHC alleles identified for the data instance i have a value of 1. Otherwise, the remaining elements have a value of 0.
  • the example is described herein with 4 identified MHC allele types, the number of MHC allele types can be hundreds or thousands in practice.
  • each data instance i typically contains at most 6 different MHC allele types in association with the peptide sequence p i .
  • the encoding module 314 also encodes the label y i for each data instance i as a binary variable having values from the set of ⁇ 0, 1 ⁇ , in which a value of 1 indicates that peptide x i was presented by one of the associated MHC alleles a i , and a value of 0 indicates that peptide x i was not presented by any of the associated MHC alleles a i .
  • the encoding module 314 may additionally scale the values using various functions, such as the log function having a range of [ ⁇ , ⁇ ] for ion current values between [0, ⁇ ].
  • the encoding module 314 may represent a pair of allele-interacting variables x h i for peptide p i and an associated MHC allele h as a row vector in which numerical representations of allele-interacting variables are concatenated one after the other.
  • the encoding module 314 may represent x h i as a row vector equal to [p i ], [p i b h i ], [p i s h i ], or [p i b h i s h i ], where b h i is the binding affinity prediction for peptide p i and associated MHC allele h, and similarly for s h i for stability.
  • one or more combination of allele-interacting variables may be stored individually (e.g., as individual vectors or matrices).
  • the encoding module 314 represents binding affinity information by incorporating measured or predicted values for binding affinity in the allele-interacting variables x h i .
  • the encoding module 314 represents binding stability information by incorporating measured or predicted values for binding stability in the allele-interacting variables x h i ,
  • the encoding module 314 represents binding on-rate information by incorporating measured or predicted values for binding on-rate in the allele-interacting variables x h i .
  • the vector T k can be included in the allele-interacting variables x h i .
  • the encoding module 314 represents RNA expression information of MHC alleles by incorporating RNA-seq based expression levels of MHC alleles in the allele-interacting variables x h i .
  • the encoding module 314 may represent the allele-noninteracting variables w i as a row vector in which numerical representations of allele-noninteracting variables are concatenated one after the other.
  • w i may be a row vector equal to [c i ] or [c i m i w i ] in which w i is a row vector representing any other allele-noninteracting variables in addition to the C-terminal flanking sequence of peptide p i and the mRNA quantification measurement m i associated with the peptide.
  • one or more combination of allele-noninteracting variables may be stored individually (e.g., as individual vectors or matrices).
  • the encoding module 314 represents turnover rate of source protein for a peptide sequence by incorporating the turnover rate or half-life in the allele-noninteracting variables w i .
  • the encoding module 314 represents length of source protein or isoform by incorporating the protein length in the allele-noninteracting variables w i .
  • the encoding module 314 represents activation of immunoproteasome by incorporating the mean expression of the immunoproteasome-specific proteasome subunits including the ⁇ 1 i , ⁇ 2 i , ⁇ 5 i subunits in the allele-noninteracting variables w i .
  • the encoding module 314 represents the RNA-seq abundance of the source protein of the peptide or gene or transcript of a peptide (quantified in units of FPKM, TPM by techniques such as RSEM) can be incorporating the abundance of the source protein in the allele-noninteracting variables w i .
  • the encoding module 314 represents the probability that the transcript of origin of a peptide will undergo nonsense-mediated decay (NMD) as estimated by the model in, for example, Rivas et. al. Science, 2015 by incorporating this probability in the allele-noninteracting variables w i .
  • NMD nonsense-mediated decay
  • the encoding module 314 represents the activation status of a gene module or pathway assessed via RNA-seq by, for example, quantifying expression of the genes in the pathway in units of TPM using e.g., RSEM for each of the genes in the pathway then computing a summary statistics, e.g., the mean, across genes in the pathway.
  • the mean can be incorporated in the allele-noninteracting variables w i .
  • the encoding module 314 represents the copy number of the source gene by incorporating the copy number in the allele-noninteracting variables w i .
  • the encoding module 314 represents the TAP binding affinity by including the measured or predicted TAP binding affinity (e.g., in nanomolar units) in the allele-noninteracting variables w i .
  • the encoding module 314 represents TAP expression levels by including TAP expression levels measured by RNA-seq (and quantified in units of TPM by e.g., RSEM) in the allele-noninteracting variables w i .
  • the encoding module 314 represents tumor type as a length-one one-hot encoded vector over the alphabet of tumor types (e.g., NSCLC, melanoma, colorectal cancer, etc). These one-hot-encoded variables can be included in the allele-noninteracting variables w i .
  • the encoding module 314 represents MHC allele suffixes by treating 4-digit HLA alleles with different suffixes.
  • HLA-A*24:09N is considered a different allele from HLA-A*24:09 for the purpose of the model.
  • the probability of presentation by an N-suffixed MHC allele can be set to zero for all peptides, because HLA alleles ending in the N suffix are not expressed.
  • the encoding module 314 represents tumor subtype as a length-one one-hot encoded vector over the alphabet of tumor subtypes (e.g., lung adenocarcinoma, lung squamous cell carcinoma, etc). These onehot-encoded variables can be included in the allele-noninteracting variables w i .
  • smoking history can be encoded as a length-one one-hot-encoded variable over an alphabet of smoking severity. For example, smoking status can be rated on a 1-5 scale, where 1 indicates nonsmokers, and 5 indicates current heavy smokers. Because smoking history is primarily relevant to lung tumors, when training a model on multiple tumor types, this variable can also be defined to be equal to 1 if the patient has a history of smoking and the tumor type is lung tumors and zero otherwise.
  • the encoding module 314 represents distribution of expression levels of a particular gene or transcript for each gene or transcript in the human genome as summary statistics (e.g., mean, median) of distribution of expression levels by using reference databases such as TCGA.
  • summary statistics e.g., mean, median
  • TCGA reference databases
  • the encoding module 314 represents mutation type as a length-one one-hot-encoded variable over the alphabet of mutation types (e.g., missense, frameshift, NMD-inducing, etc). These onehot-encoded variables can be included in the allele-noninteracting variables w i .
  • the encoding module 314 represents protein-level features of protein as the value of the annotation (e.g., 5′ UTR length) of the source protein in the allele-noninteracting variables w i .
  • the encoding module 314 represents residue-level annotations of the source protein for peptide p k by including an indicator variable, that is equal to 1 if peptide p k overlaps with a helix motif and 0 otherwise, or that is equal to 1 if peptide p k is completely contained with within a helix motif in the allele-noninteracting variables w i .
  • a feature representing proportion of residues in peptide p k that are contained within a helix motif annotation can be included in the allele-noninteracting variables w i .
  • the encoding module 314 represents type of proteins or isoforms in the human proteome as an indicator vector o k that has a length equal to the number of proteins or isoforms in the human proteome, and the corresponding element o k i is 1 if peptide p k comes from protein i and 0 otherwise.
  • the encoding module 314 may also represent the overall set of variables z i for peptide p i and an associated MHC allele h as a row vector in which numerical representations of the allele-interacting variables x i and the allele-noninteracting variables w i are concatenated one after the other.
  • the encoding module 314 may represent z h i as a row vector equal to [x h i w i ] or [w i x h i ].
  • the training module 316 constructs one or more presentation models that generate likelihoods of whether peptide sequences will be presented by MHC alleles associated with the peptide sequences. Specifically, given a peptide sequence p k and a set of MHC alleles a k associated with the peptide sequence p k , each presentation model generates an estimate u k indicating a likelihood that the peptide sequence p k will be presented by one or more of the associated MHC alleles a k .
  • the training module 316 constructs the one more presentation models based on the training data sets stored in store 170 generated from the presentation information stored in 165 .
  • all of the presentation models capture the dependence between independent variables and dependent variables in the training data 170 such that a loss function is minimized.
  • the loss function l(y i ⁇ S , u i ⁇ S ; ⁇ ) represents discrepancies between values of dependent variables y i ⁇ S for one or more data instances Sin the training data 170 and the estimated likelihoods u i ⁇ S for the data instances S generated by the presentation model.
  • the loss function (y i ⁇ S , u i ⁇ S ; ⁇ ) is the negative log likelihood function given by equation (1a) as follows:
  • the loss function is the mean squared loss given by equation 1b as follows:
  • the presentation model may be a parametric model in which one or more parameters ⁇ mathematically specify the dependence between the independent variables and dependent variables.
  • various parameters of parametric-type presentation models that minimize the loss function are determined through gradient-based numerical optimization algorithms, such as batch gradient algorithms, stochastic gradient algorithms, and the like.
  • the presentation model may be a non-parametric model in which the model structure is determined from the training data 170 and is not strictly based on a fixed set of parameters.
  • the training module 316 may construct the presentation models to predict presentation likelihoods of peptides on a per-allele basis. In this case, the training module 316 may train the presentation models based on data instances S in the training data 170 generated from cells expressing single MHC alleles.
  • the training module 316 models the estimated presentation likelihood u k for peptide p k for a specific allele h by:
  • ⁇ ( ⁇ ) is any function, and is herein throughout is referred to as a transformation function for convenience of description.
  • g h ( ⁇ ) is any function, is herein throughout referred to as a dependency function for convenience of description, and generates dependency scores for the allele-interacting variables x h k based on a set of parameters ⁇ h determined for MHC allele h.
  • the values for the set of parameters ⁇ h for each MHC allele h can be determined by minimizing the loss function with respect to ⁇ h , where i is each instance in the subset S of training data 170 generated from cells expressing the single MHC allele h.
  • the output of the dependency function g h (x h k ; ⁇ h ) represents a dependency score for the MHC allele h indicating whether the MHC allele h will present the corresponding neoantigen based on at least the allele interacting features x h k , and in particular, based on positions of amino acids of the peptide sequence of peptide p k .
  • the dependency score for the MHC allele h may have a high value if the MHC allele h is likely to present the peptide p k , and may have a low value if presentation is not likely.
  • the transformation function ⁇ ( ⁇ ) transforms the input, and more specifically, transforms the dependency score generated by g h (x h k ; ⁇ h ) in this case, to an appropriate value to indicate the likelihood that the peptide p k will be presented by an MHC allele.
  • ⁇ ( ⁇ ) is a function having the range within [0, 1] for an appropriate domain range.
  • ⁇ ( ⁇ ) is the expit function given by:
  • ⁇ ( ⁇ ) can also be the hyperbolic tangent function given by:
  • ⁇ ( ⁇ ) can be any function such as the identity function, the exponential function, the log function, and the like.
  • the per-allele likelihood that a peptide sequence p k will be presented by a MHC allele h can be generated by applying the dependency function g h ( ⁇ ) for the MHC allele h to the encoded version of the peptide sequence p k to generate the corresponding dependency score.
  • the dependency score may be transformed by the transformation function ⁇ ( ⁇ ) to generate a per-allele likelihood that the peptide sequence p k will be presented by the MHC allele h.
  • the dependency function g h ( ⁇ ) is an affine function given by:
  • the dependency function g h ( ⁇ ) is a network function given by:
  • network models are advantageous because the presentation model can incorporate non-linearity and process data having different lengths of amino acid sequences. Specifically, through non-linear modeling, network models can capture interaction between amino acids at different positions in a peptide sequence and how this interaction affects peptide presentation.
  • network models NN h ( ⁇ ) may be structured as feed-forward networks, such as artificial neural networks (ANN), convolutional neural networks (CNN), deep neural networks (DNN), and/or recurrent networks, such as long short-term memory networks (LSTM), bi-directional recurrent networks, deep bi-directional recurrent networks, and the like.
  • ANN artificial neural networks
  • CNN convolutional neural networks
  • DNN deep neural networks
  • recurrent networks such as long short-term memory networks (LSTM), bi-directional recurrent networks, deep bi-directional recurrent networks, and the like.
  • NN h ( ⁇ ) denotes the output(s) from a network model associated with MHC allele h.
  • the network model NN 3 ( ⁇ ) is associated with a set of ten parameters ⁇ 3 (1), ⁇ 3 (2), . . . , ⁇ 3 (10).
  • the set of parameters ⁇ h may correspond to a set of parameters for the single network model, and thus, the set of parameters ⁇ h may be shared by all MHC alleles.
  • the network model NN H ( ⁇ ) includes m output nodes each corresponding to an MHC allele.
  • the single network model NN H ( ⁇ ) may be a network model that outputs a dependency score given the allele interacting variables x h k and the encoded protein sequence d h of an MHC allele h.
  • the set of parameters ⁇ h may again correspond to a set of parameters for the single network model, and thus, the set of parameters ⁇ h may be shared by all MHC alleles.
  • NN h ( ⁇ ) may denote the output of the single network model NN H ( ⁇ ) given inputs [x h k d h ] to the single network model.
  • Such a network model is advantageous because peptide presentation probabilities for MHC alleles that were unknown in the training data can be predicted just by identification of their protein sequence.
  • FIG. 6B illustrates an example network model NN H ( ⁇ ) shared by MHC alleles.
  • dependency function g h ( ⁇ ) can be expressed as:
  • g′ h (x h k ; ⁇ ′ h ) is the affine function with a set of parameters ⁇ ′ h , the network function, or the like, with a bias parameter ⁇ h 0 in the set of parameters for allele interacting variables for the MHC allele that represents a baseline probability of presentation for the MHC allele h.
  • the bias parameter ⁇ h 0 may be shared according to the gene family of the MHC allele h. That is, the bias parameter ⁇ h 0 for MHC allele h may be equal to ⁇ gene(h) 0 , where gene(h) is the gene family of MHC allele h. For example, MHC alleles HLA-A*02:01, HLA-A*02:02, and HLA-A*02:03 may be assigned to the gene family of “HLA-A,” and the bias parameter ⁇ h 0 for each of these MHC alleles may be shared.
  • the training module 316 incorporates allele-noninteracting variables and models the estimated presentation likelihood u k for peptide p k by:
  • w k denotes the encoded allele-noninteracting variables for peptide p k
  • g w ( ⁇ ) is a function for the allele-noninteracting variables w k based on a set of parameters ⁇ w determined for the allele-noninteracting variables.
  • the values for the set of parameters ⁇ h for each MHC allele h and the set of parameters ⁇ w for allele-noninteracting variables can be determined by minimizing the loss function with respect to ⁇ h and ⁇ w , where i is each instance in the subset S of training data 170 generated from cells expressing single MHC alleles.
  • the output of the dependency function g w (w k ; ⁇ w ) represents a dependency score for the allele noninteracting variables indicating whether the peptide p k will be presented by one or more MHC alleles based on the impact of allele noninteracting variables.
  • the dependency score for the allele noninteracting variables may have a high value if the peptide p k is associated with a C-terminal flanking sequence that is known to positively impact presentation of the peptide p k , and may have a low value if the peptide p k is associated with a C-terminal flanking sequence that is known to negatively impact presentation of the peptide p k .
  • the per-allele likelihood that a peptide sequence p k will be presented by a MHC allele h can be generated by applying the function g h ( ⁇ ) for the MHC allele h to the encoded version of the peptide sequence p k to generate the corresponding dependency score for allele interacting variables.
  • the function g w ( ⁇ ) for the allele noninteracting variables are also applied to the encoded version of the allele noninteracting variables to generate the dependency score for the allele noninteracting variables. Both scores are combined, and the combined score is transformed by the transformation function ⁇ ( ⁇ ) to generate a per-allele likelihood that the peptide sequence p k will be presented by the MHC allele h.
  • the training module 316 may include allele-noninteracting variables w k in the prediction by adding the allele-noninteracting variables w k to the allele-interacting variables x h k in equation (2).
  • the presentation likelihood can be given by:
  • the dependency function g w ( ⁇ ) for allele noninteracting variables may be an affine function or a network function in which a separate network model is associated with allele-noninteracting variables w k .
  • the dependency function g w ( ⁇ ) is an affine function given by:
  • the dependency function g w ( ⁇ ) may also be a network function given by:
  • NN w ( ⁇ ) having an associated parameter in the set of parameters ⁇ w .
  • g′ w (w k ; ⁇ ′ w ) is the affine function, the network function with the set of allele noninteracting parameters ⁇ ′ w , or the like
  • m k is the mRNA quantification measurement for peptide p k
  • h( ⁇ ) is a function transforming the quantification measurement
  • ⁇ w m is a parameter in the set of parameters for allele noninteracting variables that is combined with the mRNA quantification measurement to generate a dependency score for the mRNA quantification measurement.
  • h( ⁇ ) is the log function, however in practice h( ⁇ ) may be any one of a variety of different functions.
  • the dependency function the dependency function g w ( ⁇ ) for the allele-noninteracting variables can be given by:
  • g′ w (w k ; ⁇ ′ w ) is the affine function, the network function with the set of allele noninteracting parameters ⁇ ′ w , or the like
  • o k is the indicator vector described above representing proteins and isoforms in the human proteome for peptide p k
  • ⁇ w o is a Set of parameters in the set of parameters for allele noninteracting variables that is combined with the indicator vector.
  • a parameter regularization term such as ⁇ w o ⁇ , where ⁇ represents L1 norm, L2 norm, a combination, or the like, can be added to the loss function when determining the value of the parameters.
  • the optimal value of the hyperparameter ⁇ can be determined through appropriate methods.
  • u k 3 ⁇ ( w k ⁇ w +x 3 k ⁇ 3 ),
  • w k are the identified allele-noninteracting variables for peptide p k
  • ⁇ w are the set of parameters determined for the allele-noninteracting variables.
  • u k 3 ⁇ ( NN w ( w k ; ⁇ w )+ NN 3 ( x 3 k ; ⁇ 3 ))
  • w k are the identified allele-interacting variables for peptide p k
  • ⁇ w are the set of parameters determined for allele-noninteracting variables.
  • the network model NN w ( ⁇ ) receives the allele-noninteracting variables w k for peptide p k and generates the output NN w (w k ).
  • the outputs are combined and mapped by function ⁇ ( ⁇ ) to generate the estimated presentation likelihood u k .
  • the training module 316 may also construct the presentation models to predict presentation likelihoods of peptides in a multiple-allele setting where two or more MHC alleles are present. In this case, the training module 316 may train the presentation models based on data instances Sin the training data 170 generated from cells expressing single MHC alleles, cells expressing multiple MHC alleles, or a combination thereof.
  • the training module 316 models the estimated presentation likelihood u k for peptide p k in association with a set of multiple MHC alleles H as a function of the presentation likelihoods u k h ⁇ H determined for each of the MHC alleles h in the set H determined based on cells expressing single-alleles, as described above in conjunction with equations (2)-(11).
  • the presentation likelihood u k can be any function of u k h ⁇ H .
  • the function is the maximum function, and the presentation likelihood u k can be determined as the maximum of the presentation likelihoods for each MHC allele h in the set H.
  • the training module 316 models the estimated presentation likelihood u k for peptide p k by:
  • elements a h k are 1 for the multiple MHC alleles H associated with peptide sequence p k and x h k denotes the encoded allele-interacting variables for peptide p k and the corresponding MHC alleles.
  • the values for the set of parameters ⁇ h for each MHC allele h can be determined by minimizing the loss function with respect to ⁇ h , where i is each instance in the subset S of training data 170 generated from cells expressing single MHC alleles and/or cells expressing multiple MHC alleles.
  • the dependency function g h may be in the form of any of the dependency functions g h introduced above in sections VIII.B.1.
  • the presentation likelihood that a peptide sequence p k will be presented by one or more MHC alleles h can be generated by applying the dependency function g h ( ⁇ ) to the encoded version of the peptide sequence p k for each of the MHC alleles H to generate the corresponding score for the allele interacting variables.
  • the scores for each MHC allele h are combined, and transformed by the transformation function ⁇ ( ⁇ ) to generate the presentation likelihood that peptide sequence p k will be presented by the set of MHC alleles H.
  • the presentation model of equation (13) is different from the per-allele model of equation (2), in that the number of associated alleles for each peptide p k can be greater than 1. In other words, more than one element in a h k can have values of 1 for the multiple MHC alleles H associated with peptide sequence p k .
  • u k ⁇ ( x 2 k ⁇ 2 +x 3 k ⁇ 3 ),
  • u k ⁇ ( NN 2 ( x 2 k ; ⁇ 2 )+ NN 3 ( x 3 k ; ⁇ 3 )),
  • the outputs are combined and mapped by function ⁇ ( ⁇ ) to generate the estimated presentation likelihood u k .
  • the training module 316 incorporates allele-noninteracting variables and models the estimated presentation likelihood u k for peptide p k by:
  • w k denotes the encoded allele-noninteracting variables for peptide p k .
  • the values for the set of parameters ⁇ h for each MHC allele h and the set of parameters ⁇ w for allele-noninteracting variables can be determined by minimizing the loss function with respect to ⁇ h and ⁇ w , where i is each instance in the subset S of training data 170 generated from cells expressing single MHC alleles and/or cells expressing multiple MHC alleles.
  • the dependency function g w may be in the form of any of the dependency functions g w introduced above in sections VIII.B.3.
  • the presentation likelihood that a peptide sequence p k will be presented by one or more MHC alleles H can be generated by applying the function g h ( ⁇ ) to the encoded version of the peptide sequence p k for each of the MHC alleles H to generate the corresponding dependency score for allele interacting variables for each MHC allele h.
  • the function g w ( ⁇ ) for the allele noninteracting variables is also applied to the encoded version of the allele noninteracting variables to generate the dependency score for the allele noninteracting variables.
  • the scores are combined, and the combined score is transformed by the transformation function ⁇ ( ⁇ ) to generate the presentation likelihood that peptide sequence p k will be presented by the MHC alleles H.
  • the number of associated alleles for each peptide p k can be greater than 1. In other words, more than one element in a h k can have values of 1 for the multiple MHC alleles H associated with peptide sequence p k .
  • u k ⁇ ( w k ⁇ w +x 2 k +x 3 k ⁇ 3 ),
  • w k are the identified allele-noninteracting variables for peptide p k
  • ⁇ w are the set of parameters determined for the allele-noninteracting variables.
  • u k ⁇ ( NN w ( w k ; ⁇ w )+ NN 2 ( x 2 k ; ⁇ 2 )+ NN 3 ( x k 3 ; ⁇ 3 ))
  • w k are the identified allele-interacting variables for peptide p k
  • ⁇ w are the set of parameters determined for allele-noninteracting variables.
  • the network model NN w ( ⁇ ) receives the allele-noninteracting variables w k for peptide p k and generates the output NN w (w k ). The outputs are combined and mapped by function ⁇ ( ⁇ ) to generate the estimated presentation likelihood u k .
  • the training module 316 may include allele-noninteracting variables w k in the prediction by adding the allele-noninteracting variables w k to the allele-interacting variables x h k in equation (15).
  • the presentation likelihood can be given by:
  • the training module 316 models the estimated presentation likelihood u k for peptide p k by:
  • vector v is a vector in which element v h corresponds to a h k ⁇ u′ k h
  • s( ⁇ ) is a function mapping the elements of v
  • r( ⁇ ) is a clipping function that clips the value of the input into a given range.
  • s( ⁇ ) may be the summation function or the second-order function, but it is appreciated that in other embodiments, s( ⁇ ) can be any function such as the maximum function.
  • the values for the set of parameters ⁇ for the implicit per-allele likelihoods can be determined by minimizing the loss function with respect to ⁇ , where i is each instance in the subset S of training data 170 generated from cells expressing single MHC alleles and/or cells expressing multiple MHC alleles.
  • the presentation likelihood in the presentation model of equation (17) is modeled as a function of implicit per-allele presentation likelihoods u′ k h that each correspond to the likelihood peptide p k will be presented by an individual MHC allele h.
  • the implicit per-allele likelihood is distinct from the per-allele presentation likelihood of section VIII.B in that the parameters for implicit per-allele likelihoods can be learned from multiple allele settings, in which direct association between a presented peptide and the corresponding MHC allele is unknown, in addition to single-allele settings.
  • the presentation model can estimate not only whether peptide p k will be presented by a set of MHC alleles H as a whole, but can also provide individual likelihoods u′ k h ⁇ H that indicate which MHC allele h most likely presented peptide p k .
  • An advantage of this is that the presentation model can generate the implicit likelihoods without training data for cells expressing single MHC alleles.
  • r( ⁇ ) is a function having the range [0, 1].
  • r( ⁇ ) may be the clip function:
  • r( ⁇ ) is the hyperbolic tangent function given by:
  • s( ⁇ ) is a summation function
  • the presentation likelihood is given by summing the implicit per-allele presentation likelihoods:
  • the implicit per-allele presentation likelihood for MHC allele h is generated by:
  • the presentation likelihood that a peptide sequence p k will be presented by one or more MHC alleles H can be generated by applying the function g h ( ⁇ ) to the encoded version of the peptide sequence p k for each of the MHC alleles H to generate the corresponding dependency score for allele interacting variables.
  • Each dependency score is first transformed by the function ⁇ ( ⁇ ) to generate implicit per-allele presentation likelihoods u′ k h .
  • the per-allele likelihoods u′ k h are combined, and the clipping function may be applied to the combined likelihoods to clip the values into a range [0, 1] to generate the presentation likelihood that peptide sequence p k will be presented by the set of MHC alleles H
  • the dependency function g h may be in the form of any of the dependency functions g h introduced above in sections VIII.B.1.
  • u k r ( ⁇ ( NN 2 ( x 2 k ; ⁇ 2 ))+ ⁇ ( NN 3 ( x 3 k ; ⁇ 3 ))),
  • Each output is mapped by function ⁇ ( ⁇ ) and combined to generate the estimated presentation likelihood u k .
  • r( ⁇ ) is the log function and ⁇ ( ⁇ ) is the exponential function.
  • the implicit per-allele presentation likelihood for MHC allele h is generated by:
  • the presentation likelihood that a peptide sequence p k will be presented by one or more MHC alleles H can be generated by applying the function g h ( ⁇ ) to the encoded version of the peptide sequence p k for each of the MHC alleles H to generate the corresponding dependency score for allele interacting variables for each MHC allele h.
  • the function g w ( ⁇ ) for the allele noninteracting variables is also applied to the encoded version of the allele noninteracting variables to generate the dependency score for the allele noninteracting variables.
  • the score for the allele noninteracting variables are combined to each of the dependency scores for the allele interacting variables.
  • Each of the combined scores are transformed by the function ⁇ ( ⁇ ) to generate the implicit per-allele presentation likelihoods.
  • the implicit likelihoods are combined, and the clipping function may be applied to the combined outputs to clip the values into a range [0,1] to generate the presentation likelihood that peptide sequence p k will be presented by the MHC alleles H
  • the dependency function g w may be in the form of any of the dependency functions g w introduced above in sections VIII.B.3.
  • u k r ( ⁇ ( w k ⁇ w +x 2 k ⁇ 2 )+ ⁇ ( w k ⁇ w +x 3 k ⁇ 3 )),
  • w k are the identified allele-noninteracting variables for peptide p k
  • ⁇ w are the set of parameters determined for the allele-noninteracting variables.
  • u k r ( ⁇ ( NN w ( w k ; ⁇ w )+ NN 2 ( x 2 k ; ⁇ 2 ))+ ⁇ ( NN w ( w k ; ⁇ w )+ NN 3 ( x 3 k ; ⁇ 3 )))
  • w k are the identified allele-interacting variables for peptide p k
  • ⁇ w are the set of parameters determined for allele-noninteracting variables.
  • the network model NN w ( ⁇ ) receives the allele-noninteracting variables w k for peptide p k and generates the output NN w (w k ).
  • the outputs are combined and mapped by function ⁇ ( ⁇ ).
  • the implicit per-allele presentation likelihood for MHC allele h is generated by:
  • s( ⁇ ) is a second-order function
  • the estimated presentation likelihood u k for peptide p k is given by:
  • elements u′ k h are the implicit per-allele presentation likelihood for MHC allele h.
  • the values for the set of parameters ⁇ for the implicit per-allele likelihoods can be determined by minimizing the loss function with respect to ⁇ , where i is each instance in the subset S of training data 170 generated from cells expressing single MHC alleles and/or cells expressing multiple MHC alleles.
  • the implicit per-allele presentation likelihoods may be in any form shown in equations (18), (20), and (22) described above.
  • the model of equation (23) may imply that there exists a possibility peptide p k will be presented by two MHC alleles simultaneously, in which the presentation by two HLA alleles is statistically independent.
  • the presentation likelihood that a peptide sequence p k will be presented by one or more MHC alleles H can be generated by combining the implicit per-allele presentation likelihoods and subtracting the likelihood that each pair of MHC alleles will simultaneously present the peptide p k from the summation to generate the presentation likelihood that peptide sequence p k will be presented by the MHC alleles H.
  • u k ⁇ ( x 2 k ⁇ 2 )+ ⁇ ( x 3 k ⁇ 3 ) ⁇ ( x 2 k ⁇ 2 ) ⁇ ( x 3 k ⁇ 3 ),
  • u k ⁇ ( NN 2 ( x 2 k ; ⁇ 2 ))+ ⁇ ( NN 3 ( x 3 k ; ⁇ 3 )) ⁇ ( NN 2 ( x 2 k ; ⁇ 2 )) ⁇ ( NN 3 ( x 3 k ; ⁇ 3 )),
  • the prediction module 320 receives sequence data and selects candidate neoantigens in the sequence data using the presentation models.
  • the sequence data may be DNA sequences, RNA sequences, and/or protein sequences extracted from tumor tissue cells of patients.
  • the prediction module 320 processes the sequence data into a plurality of peptide sequences p k having 8-15 amino acids.
  • the prediction module 320 may process the given sequence “IEFROEIFJEF (SEQ ID NO: 15) into three peptide sequences having 9 amino acids “IEFROEIFJ (SEQ ID NO: 16),” “EFROEIFJE (SEQ ID NO: 17),” and “FROEIFJEF (SEQ ID NO: 18).”
  • the prediction module 320 may identify candidate neoantigens that are mutated peptide sequences by comparing sequence data extracted from normal tissue cells of a patient with the sequence data extracted from tumor tissue cells of the patient to identify portions containing one or more mutations.
  • the presentation module 320 applies one or more of the presentation models to the processed peptide sequences to estimate presentation likelihoods of the peptide sequences.
  • the prediction module 320 may select one or more candidate neoantigen peptide sequences that are likely to be presented on tumor HLA molecules by applying the presentation models to the candidate neoantigens.
  • the presentation module 320 selects candidate neoantigen sequences that have estimated presentation likelihoods above a predetermined threshold.
  • the presentation model selects the N candidate neoantigen sequences that have the highest estimated presentation likelihoods (where N is generally the maximum number of epitopes that can be delivered in a vaccine).
  • a vaccine including the selected candidate neoantigens for a given patient can be injected into the patient to induce immune responses.
  • test data T were subsets of training data 170 that were not used to train the presentation models or a separate dataset from the training data 170 that have similar variables and data structures as the training data 170 .
  • a relevant metric indicative of the performance of a presentation models is:
  • a peptide p i in the test data T was predicted to be presented on one or more associated HLA alleles if the corresponding likelihood estimate u i is greater or equal to a given threshold value t.
  • Another relevant metric indicative of the performance of presentation models is:
  • AUC area-under-curve
  • FPR false positive rate
  • FIG. 13A compares performance results of an example presentation model, as presented herein, and state-of-the-art models for predicting peptide presentation on multiple-allele mass spectrometry data. Results showed that the example presentation model performed significantly better at predicting peptide presentation than state-of-the-art models based on affinity and stability predictions.
  • the example presentation model shown in FIG. 13A as “MS” was the maximum of per-alleles presentation model shown in equation (12), using the affine dependency function g h ( ⁇ ) and the expit function ⁇ ( ⁇ ).
  • the example presentation model was trained based on a subset of the single-allele HLA-A*02:01 mass spectrometry data from the IEDB data set (data set “D1”) (data can be found at http://www.iedb.org/doc/mhc_ligand_full.zip) and a subset of the single-allele HLA-B*07:02 mass spectrometry from the IEDB data set (data set “D2”) (data can be found at http://www.iedb.org/doc/mhc_ligand_full.zip). All peptides from source protein that contain presented peptides in the test set were eliminated from the training data such that the example presentation model could not simply memorize the sequences of presented antigens.
  • the model shown in FIG. 13A as “Affinity” was a model similar to the current state-of-the-art model that predicts peptide presentation based on affinity predictions NETMHCpan.
  • NETMHCpan is provided in detail at http://www.cbs.dtu.dk/services/NetMHCpan/.
  • Stability was a model similar to the current state-of-the-art model that predicts peptide presentation based on stability predictions NETMHCstab.
  • Implementation of NETMHCstab is provided in detail at http://www.cbs.dtu.dk/services/NetMHCstab-1.0/.
  • test data that is a subset of the multiple-allele JY cell line HLA-A*02:01 and HLA-B*07:02 mass spectrometry data from the Bassani-Sternberg data set (data set “D3”) (data can be found at www.ebi.ac.uk/pride/archive/projects/PXD000394).
  • the error bars (as indicated in solid lines) show 95% confidence intervals.
  • the example presentation model trained on mass spectrometry data had a significantly higher PPV value at 10% recall rate relative to the state-of-the-art models that predict peptide presentation based on MHC binding affinity predictions or MHC binding stability predictions.
  • the example presentation model had approximately 14% higher PPV than the model based on affinity predictions, and had approximately 12% higher PPV than the model based on stability predictions.
  • FIG. 13B compares performance results of another example presentation model, as presented herein, and state-of-the-art models for predicting peptide presentation on T-cell epitope data.
  • T-cell epitope data contains peptide sequences that were presented by MHC alleles on the cell surface, and recognized by T-cells. Results showed that even though the example presentation model is trained based on mass spectrometry data, the example presentation model performed significantly better at predicting T-cell epitopes than state-of-the-art models based on affinity and stability predictions. In other words, the results of FIG.
  • the example presentation model shown in FIG. 13B as “MS” was the per-allele presentation model shown in equation (2), using the affine transformation function g h ( ⁇ ) and the expit function ⁇ ( ⁇ ) that was trained based on a subset of data set D1. All peptides from source protein that contain presented peptides in the test set were eliminated from the training data such that the presentation model could not simply memorize the sequences of presented antigens.
  • Each of the models were applied to the test data that is a subset of mass spectrometry data on HLA-A*02:01 T-cell epitope data (data set “D4”) (data can be found at www.iedb.org/doc/tcell full v3.zip).
  • the model shown in FIG. 13B as “Affinity” was a model similar to the current state-of-the-art model that predicts peptide presentation based on affinity predictions NETMHCpan
  • Stability was a model similar to the current state-of-the-art model that predicts peptide presentation based on stability predictions NETMHCstab.
  • the error bars (as indicated in solid lines) show 95% confidence intervals.
  • the per-allele presentation model trained on mass spectrometry data had a significantly higher PPV value at 10% recall rate than the state-of-the-art models that predict peptide presentation based on MHC binding affinity or MHC binding stability predictions even though the presentation model was not trained based on protein sequences that contained presented peptides.
  • the per-allele presentation model had approximately 9% higher PPV than the model based on affinity predictions, and had approximately 8% higher PPV than the model based on stability predictions.
  • FIG. 13C compares performance results for an example function-of-sums model (equation (13)), an example sum-of-functions model (equation (19)), and an example second order model (equation (23)) for predicting peptide presentation on multiple-allele mass spectrometry data. Results showed that the sum-of-functions model and second order model performed better than the function-of-sums model. This is because the function-of-sums model implies that alleles in a multiple-allele setting can interfere with each other for peptide presentation, when in reality, the presentation of peptides are effectively independent.
  • the example presentation model labeled as “sigmoid-of-sums” in FIG. 13C was the function-of-sums model using a network dependency function g h ( ⁇ ) the identity function ⁇ ( ⁇ ), and the expit function r( ⁇ ).
  • the example model labeled as “sum-of-sigmoids” was the sum-of-functions model in equation (19) with a network dependency function g h ( ⁇ ), the expit function ⁇ ( ⁇ ), and the identity function r( ⁇ ).
  • the example model labeled as “hyperbolic tangent” was the sum-of-functions model in equation (19) with a network dependency function g h ( ⁇ ), the expit function ⁇ ( ⁇ ), and the hyperbolic tangent function r( ⁇ ).
  • the example model labeled as “second order” was the second order model in equation (23) using an implicit per-allele presentation likelihood form shown in equation (18) with a network dependency function g h ( ⁇ ), and the expit function ⁇ ( ⁇ ).
  • Each model was trained based on a subset of data set D1, D2, and D3.
  • the example presentation models were applied to a test data that is a random subset of data set D3 that did not overlap with the training data.
  • the first column refers to the AUC of the ROC when each presentation model was applied to the test set
  • the second column refers to the value of the negative log likelihood loss
  • the third column refers to the PPV at 10% recall rate.
  • the performance of presentation models “sum-of-sigmoids,” “hyperbolic tangent,” and “second order” were approximately tied at approximately 15-16% PPV at 10% recall, while the performance of the model “sigmoid-of-sums” was slightly lower at approximately 11%.
  • FIG. 13D compares performance results for two example presentation models that are trained with and without single-allele mass spectrometry data on predicting peptide presentation for multiple-allele mass spectrometry data. The results indicated that example presentation models that are trained without single-allele data achieve comparable performance to that of example presentation models trained with single-allele data.
  • the example model “with A2/B7 single-allele data” was the “sum-of-sigmoids” presentation model in equation (19) with a network dependency function g h ( ⁇ ), the expit function ⁇ ( ⁇ ), and the identity function r( ⁇ ).
  • the model was trained based on a subset of data set D3 and single-allele mass spectrometry data for a variety of MHC alleles from the IEDB database (data can be found at: http://www.iedb.org/doc/mhc_ligand_full.zip).
  • the example model “without A2/B7 single-allele data” was the same model, but trained based on a subset of the multiple-allele D3 data set without single-allele mass spectrometry data for alleles HLA-A*02:01 and HLA-B*07:02, but with single-allele mass spectrometry data for other alleles.
  • cell line HCC1937 expressed HLA-B*07:02 but not HLA-A*02:01
  • cell line HCT116 expressed HLA-A*02:01 but not HLA-B*07:02.
  • the example presentation models were applied to a test data that was a random subset of data set D3 and did not overlap with the training data.
  • the column “Correlation” refers to the correlation between the actual labels that indicate whether the peptide was presented on the corresponding allele in the test data, and the label for prediction.
  • the predictions based on the implicit per-allele presentation likelihoods for MHC allele HLA-A*02:01 performed significantly better on single-allele test data for MHC allele HLA-A*02:01 rather than for MHC allele HLA-B*07:02. Similar results are shown for MHC allele HLA-B*07:02.
  • FIG. 13E shows performance for the “without A2/B7 single-allele data” and “with A2/B7 single-allele data” example models shown in FIG. 13D on single-allele mass spectrometry data for alleles HLA-A*02:01 and HLA-B*07:02 that were held out in the analysis shown in FIG. 13D .
  • Results indicate that even through the example presentation model is trained without single-allele mass spectrometry data for these two alleles, the model is able to learn binding motifs for each MHC allele.
  • “A2 model predicting B7” indicates the performance of the model when peptide presentation is predicted for single-allele HLA-B*07:02 data based on the implicit per-allele presentation likelihood estimate for MHC allele HLA-A*02:01.
  • “A2 model predicting A2” indicates the performance of the model when peptide presentation is predicted for single-allele HLA-A*02:01 based on the implicit per-allele presentation likelihood estimate for MHC allele HLA-A*02:01.
  • B7 model predicting B7 indicates the performance of the model when peptide presentation is predicted for single-allele HLA-B*07:02 data based on the implicit per-allele presentation likelihood estimate for MHC allele HLA-B*07:02.
  • B7 model predicting A2 indicates the performance of the model when peptide presentation is predicted for single-allele HLA-A*02:01 based on the implicit per-allele presentation likelihood estimate for MHC allele HLA-B*07:02.
  • the predictive capacity of implicit per-allele likelihoods for an HLA allele is significantly higher for the intended allele, and significantly lower for the other HLA allele.
  • the example presentation models correctly learned to differentiate peptide presentation of individual alleles HLA-A*02:01 and HLA-B*07:02, even though direct association between peptide presentation and these alleles were not present in the multiple-allele training data.
  • FIG. 13F shows the common anchor residues at positions 2 and 9 among nonamers predicted by the “without A2/B7 single-allele data” example model shown in FIG. 13D .
  • the peptides were predicted to be presented if the estimated likelihood was above 5%.
  • Results show that most common anchor residues in the peptides identified for presentation on the MHC alleles HLA-A*02:01 and HLA-B*07:02 matched previously known anchor motifs for these MHC alleles. This indicates that the example presentation models correctly learned peptide binding based on particular positions of amino acids of the peptide sequences, as expected.
  • amino acids L/M at position 2 and amino acids V/L at position 9 were known to be canonical anchor residue motifs (as shown in Table 4 of https://link.springer.com/article/10.1186/1745-7580-4-2) for HLA-A*02:01, and amino acid P at position 2 and amino acids L/V at position 9 were known to be canonical anchor residue motifs for HLA-B*07:02.
  • the most common anchor residue motifs at positions 2 and 9 for peptides identified the model matched the known canonical anchor residue motifs for both HLA alleles.
  • FIG. 13G compares performance results between an example presentation model that incorporated C- and N-terminal flanking sequences as allele-interacting variables, and an example presentation model that incorporated C- and N-terminal flanking sequences as allele-noninteracting variables. Results showed that incorporating C- and N-terminal flanking sequences as allele noninteracting variables significantly improved model performance. More specifically, it is valuable to identify appropriate features for peptide presentation that are common across different MHC alleles, and model them such that statistical strength for these allele-noninteracting variables are shared across MHC alleles to improve presentation model performance.
  • the example “allele-interacting” model was the sum-of-functions model using the form of implicit per-allele presentation likelihoods in equation (22) that incorporated C- and N-terminal flanking sequences as allele-interacting variables, with a network dependency function g h ( ⁇ ) and the expit function ⁇ ( ⁇ ).
  • the example “allele-noninteracting” model was the sum-of-functions model shown in equation (21) that incorporated C- and N-terminal flanking sequences as allele-noninteracting variables, with a network dependency function g h ( ⁇ ) and the expit function ⁇ ( ⁇ ).
  • the allele-noninteracting variables were modeled through a separate network dependency function g w ( ⁇ ).
  • Both models were trained on a subset of data set D3 and single-allele mass spectrometry data for a variety of MHC alleles from the IEDB database (data can be found at: http://www.iedb.org/doc/mhc_ligand_full.zip).
  • Each of the presentation models was applied to a test data set that is a random subset of data set D3 that did not overlap with the training data.
  • FIG. 13H illustrates the dependency between fraction of presented peptides for genes based on mRNA quantification for mass spectrometry data on tumor cells. Results show that there is a strong dependency between mRNA expression and peptide presentation.
  • the horizontal axis in FIG. 13G indicates mRNA expression in terms of transcripts per million (TPM) quartiles.
  • the vertical axis in FIG. 13G indicates fraction of presented epitopes from genes in corresponding mRNA expression quartiles.
  • Each solid line is a plot relating the two measurements from a tumor sample that is associated with corresponding mass spectrometry data and mRNA expression measurements.
  • FIG. 13G there is a strong positive correlation between mRNA expression, and the fraction of peptides in the corresponding gene.
  • peptides from genes in the top quartile of RNA expression are more than 20 times likely to be presented than the bottom quartile.
  • essentially 0 peptides are presented from genes that are not detected through RNA.
  • FIG. 13I shows performance of two example presentation models, one of which is trained based on mass spectrometry tumor cell data, another of which incorporates mRNA quantification data and mass spectrometry tumor cell data.
  • MHCflurry+RNA filter was a model similar to the current state-of-the-art model that predicts peptide presentation based on affinity predictions. It was implemented using MHCflurry along with a standard gene expression filter that removed all peptides from proteins with mRNA quantification measurements that were less than 3.2 FPKM. Implementation of MHCflurry is provided in detail at https://github.com/hammerlab/mhcflurry/, and at http://biorxiv.org/content/early/2017/01/22/054775.
  • Example Model, no RNA was the “sum-of-sigmoids” example presentation model shown in equation (21) with the network dependency function g h ( ⁇ ), the network dependency function g w ( ⁇ ), and the expit function ⁇ ( ⁇ ).
  • the “Example Model, no RNA” model incorporated C-terminal flanking sequences as allele-noninteracting variables through a network dependency function g w ( ⁇ ).
  • Example Model, with RNA was the “sum-of-sigmoids” presentation model shown in equation (19) with network dependency function g h ( ⁇ ), the network dependency function g w ( ⁇ ) in equation (10) incorporating mRNA quantification data through a log function, and the expit function ⁇ ( ⁇ ).
  • the “Example Model, with RNA” model incorporated C-terminal flanking sequences as allele-noninteracting variables through the network dependency functions g w ( ⁇ ) and incorporated mRNA quantification measurements through the log function.
  • Each model was trained on a combination of the single-allele mass spectrometry data from the IEDB data set, 7 cell lines from the multiple-allele mass spectrometry data from the Bassani-Sternberg data set, and 20 mass spectrometry tumor samples. Each model was applied to a test set including 5,000 held-out proteins from 7 tumor samples that constituted 9,830 presented peptides from a total of 52,156,840 peptides.
  • the “Example Model, no RNA” model has a PPV value at 20% Recall of 21%, while that of the state-of-the-art model is approximately 3%, This indicates an initial performance improvement of 18% in PPV value, even without the incorporation of mRNA quantification measurements.
  • the “Example Model, with RNA” model that incorporates mRNA quantification data into the presentation model shows a PPV value of approximately 30%, which is almost a 10% increase in performance compared to the example presentation model without mRNA quantification measurements.
  • FIG. 13J compares probability of peptide presentation for different peptide lengths between results generated by the “Example Model, with RNA” presentation model described in reference to FIG. 13I , and predicted results by state-of-the-art models that do not account for peptide length when predicting peptide presentation. Results indicated that the “Example Model, with RNA” example presentation model from FIG. 13I captured variation in likelihoods across peptides of differing lengths.
  • the horizontal axis denoted samples of peptides with lengths 8, 9, 10, and 11.
  • the vertical axis denoted the probability of peptide presentation conditioned on the lengths of the peptide.
  • the plot “Actual Test Data Probability” showed the proportion of presented peptides according to the length of the peptide in a sample test data set.
  • the presentation likelihood varied with the length of the peptide. For example, as shown in FIG. 13J , a 10 mer peptide with canonical HLA-A2 L/V anchor motifs was approximately 3 times less likely to be presented than a 9 mer with the same anchor residues.
  • the plot “Models Ignoring Length” indicated predicted measurements if state-of-the-art models that ignore peptide length were to be applied to the same test data set for presentation prediction. These models may be NetMHC versions before version 4.0, NetMHCpan versions before version 3.0, and MHCflurry, that do not take into account variation in peptide presentation according to peptide length. As shown in FIG. 13J , the proportion of presented peptides would be constant across different values of peptide length, indicating that these models would fail to capture variation in peptide presentation according to length.
  • the plot “Gritstone, with RNA” indicated measurements generated from the “Gritstone, with RNA” presentation model. As shown in FIG. 13J , the measurements generated by the “Gritstone, with RNA” model closely followed those shown in “Actual Test Data Probability” and correctly accounted for different degrees of peptide presentation for lengths 8, 9, 10, and 11.
  • u k expit(relu( x h k ⁇ W h 1 +b h 1 ) ⁇ W h 2 +b h 2 ),
  • relu( ⁇ ) is the rectified linear unit (RELU) function
  • W h i , b h 1 , W h 2 , and b h 2 are the set of parameters ⁇ determined for the model.
  • the allele interacting variables x h k consist of peptide sequences.
  • the dimensions of W h 1 are (231 ⁇ 256), the dimensions of b h 1 (1 ⁇ 256), the dimensions of W h 2 are (256 ⁇ 1), and b h 2 is a scalar.
  • values for b h 1 , b h 2 , W h 1 , and W h 2 are listed below.
  • FIG. 14 illustrates an example computer 1400 for implementing the entities shown in FIGS. 1 and 3 .
  • the computer 1400 includes at least one processor 1402 coupled to a chipset 1404 .
  • the chipset 1404 includes a memory controller hub 1420 and an input/output (I/O) controller hub 1422 .
  • a memory 1406 and a graphics adapter 1412 are coupled to the memory controller hub 1420 , and a display 1418 is coupled to the graphics adapter 1412 .
  • a storage device 1408 , an input device 1414 , and network adapter 1416 are coupled to the I/O controller hub 1422 .
  • Other embodiments of the computer 1400 have different architectures.
  • the storage device 1408 is a non-transitory computer-readable storage medium such as a hard drive, compact disk read-only memory (CD-ROM), DVD, or a solid-state memory device.
  • the memory 1406 holds instructions and data used by the processor 1402 .
  • the input interface 1414 is a touch-screen interface, a mouse, track ball, or other type of pointing device, a keyboard, or some combination thereof, and is used to input data into the computer 1400 .
  • the computer 1400 may be configured to receive input (e.g., commands) from the input interface 1414 via gestures from the user.
  • the graphics adapter 1412 displays images and other information on the display 1418 .
  • the network adapter 1416 couples the computer 1400 to one or more computer networks.
  • the computer 1400 is adapted to execute computer program modules for providing functionality described herein.
  • module refers to computer program logic used to provide the specified functionality.
  • a module can be implemented in hardware, firmware, and/or software.
  • program modules are stored on the storage device 1408 , loaded into the memory 1406 , and executed by the processor 1402 .
  • the types of computers 1400 used by the entities of FIG. 1 can vary depending upon the embodiment and the processing power required by the entity.
  • the presentation identification system 160 can run in a single computer 1400 or multiple computers 1400 communicating with each other through a network such as in a server farm.
  • the computers 1400 can lack some of the components described above, such as graphics adapters 1412 , and displays 1418 .

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Chemical & Material Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Medicinal Chemistry (AREA)
  • Public Health (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Medical Informatics (AREA)
  • Immunology (AREA)
  • Epidemiology (AREA)
  • Biotechnology (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Veterinary Medicine (AREA)
  • Pharmacology & Pharmacy (AREA)
  • Animal Behavior & Ethology (AREA)
  • Molecular Biology (AREA)
  • Theoretical Computer Science (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Biophysics (AREA)
  • Microbiology (AREA)
  • Evolutionary Biology (AREA)
  • Analytical Chemistry (AREA)
  • Hematology (AREA)
  • Oncology (AREA)
  • Proteomics, Peptides & Aminoacids (AREA)
  • Mycology (AREA)
  • General Chemical & Material Sciences (AREA)
  • Chemical Kinetics & Catalysis (AREA)
  • Biomedical Technology (AREA)
  • Urology & Nephrology (AREA)
  • Cell Biology (AREA)
  • General Physics & Mathematics (AREA)
  • Primary Health Care (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Organic Chemistry (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Food Science & Technology (AREA)
  • Bioethics (AREA)
US15/381,375 2015-12-16 2016-12-16 Neoantigen Identification, Manufacture, and Use Abandoned US20170199961A1 (en)

Priority Applications (7)

Application Number Priority Date Filing Date Title
US15/381,375 US20170199961A1 (en) 2015-12-16 2016-12-16 Neoantigen Identification, Manufacture, and Use
US15/466,729 US10055540B2 (en) 2015-12-16 2017-03-22 Neoantigen identification, manufacture, and use
US16/001,569 US10847252B2 (en) 2015-12-16 2018-06-06 Neoantigen identification, manufacture, and use
US16/040,409 US10847253B2 (en) 2015-12-16 2018-07-19 Neoantigen identification, manufacture, and use
US16/128,421 US20190034585A1 (en) 2015-12-16 2018-09-11 Neoantigen identification, manufacture, and use
US17/101,518 US11183286B2 (en) 2015-12-16 2020-11-23 Neoantigen identification, manufacture, and use
US17/101,522 US20210166784A1 (en) 2015-12-16 2020-11-23 Neoantigen identification, manufacture, and use

Applications Claiming Priority (6)

Application Number Priority Date Filing Date Title
US201562268333P 2015-12-16 2015-12-16
US201662317823P 2016-04-04 2016-04-04
US201662379986P 2016-08-26 2016-08-26
US201662394074P 2016-09-13 2016-09-13
US201662425995P 2016-11-23 2016-11-23
US15/381,375 US20170199961A1 (en) 2015-12-16 2016-12-16 Neoantigen Identification, Manufacture, and Use

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US15/466,729 Continuation US10055540B2 (en) 2015-12-16 2017-03-22 Neoantigen identification, manufacture, and use

Publications (1)

Publication Number Publication Date
US20170199961A1 true US20170199961A1 (en) 2017-07-13

Family

ID=59057557

Family Applications (7)

Application Number Title Priority Date Filing Date
US15/381,375 Abandoned US20170199961A1 (en) 2015-12-16 2016-12-16 Neoantigen Identification, Manufacture, and Use
US15/466,729 Active US10055540B2 (en) 2015-12-16 2017-03-22 Neoantigen identification, manufacture, and use
US16/001,569 Active US10847252B2 (en) 2015-12-16 2018-06-06 Neoantigen identification, manufacture, and use
US16/040,409 Active US10847253B2 (en) 2015-12-16 2018-07-19 Neoantigen identification, manufacture, and use
US16/128,421 Abandoned US20190034585A1 (en) 2015-12-16 2018-09-11 Neoantigen identification, manufacture, and use
US17/101,522 Pending US20210166784A1 (en) 2015-12-16 2020-11-23 Neoantigen identification, manufacture, and use
US17/101,518 Active US11183286B2 (en) 2015-12-16 2020-11-23 Neoantigen identification, manufacture, and use

Family Applications After (6)

Application Number Title Priority Date Filing Date
US15/466,729 Active US10055540B2 (en) 2015-12-16 2017-03-22 Neoantigen identification, manufacture, and use
US16/001,569 Active US10847252B2 (en) 2015-12-16 2018-06-06 Neoantigen identification, manufacture, and use
US16/040,409 Active US10847253B2 (en) 2015-12-16 2018-07-19 Neoantigen identification, manufacture, and use
US16/128,421 Abandoned US20190034585A1 (en) 2015-12-16 2018-09-11 Neoantigen identification, manufacture, and use
US17/101,522 Pending US20210166784A1 (en) 2015-12-16 2020-11-23 Neoantigen identification, manufacture, and use
US17/101,518 Active US11183286B2 (en) 2015-12-16 2020-11-23 Neoantigen identification, manufacture, and use

Country Status (18)

Country Link
US (7) US20170199961A1 (ja)
EP (2) EP4299136A3 (ja)
JP (2) JP7114477B2 (ja)
KR (1) KR20180107102A (ja)
CN (1) CN108601731A (ja)
AU (2) AU2016369519B2 (ja)
BR (1) BR112018012374A2 (ja)
CA (1) CA3008641A1 (ja)
CO (1) CO2018007417A2 (ja)
HK (1) HK1257865A1 (ja)
IL (2) IL259931B2 (ja)
MX (1) MX2018007204A (ja)
PE (1) PE20181535A1 (ja)
PH (1) PH12018501267A1 (ja)
RU (1) RU2729116C2 (ja)
SG (1) SG11201804957VA (ja)
TW (1) TWI765875B (ja)
WO (1) WO2017106638A1 (ja)

Cited By (34)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109712672A (zh) * 2018-12-29 2019-05-03 北京优迅医学检验实验室有限公司 检测基因重排的方法、装置、存储介质及处理器
CN109706065A (zh) * 2018-12-29 2019-05-03 深圳裕策生物科技有限公司 肿瘤新生抗原负荷检测装置及存储介质
US20190189241A1 (en) * 2016-07-20 2019-06-20 Biontech Rna Pharmaceuticals Gmbh Selecting Neoepitopes as Disease-Specific Targets for Therapy with Enhanced Efficacy
WO2019168984A1 (en) * 2018-02-27 2019-09-06 Gritstone Oncology, Inc. Neoantigen identification with pan-allele models
WO2019173687A1 (en) * 2018-03-08 2019-09-12 The Trustees Of Indiana University Constrained de novo sequencing of neo-epitope peptides using tandem mass spectrometry
WO2019200398A1 (en) * 2018-04-13 2019-10-17 Dana-Farber Cancer Institute, Inc. Ultra-sensitive detection of cancer by algorithmic analysis
US20200138923A1 (en) * 2017-06-21 2020-05-07 Transgene Personalized vaccine
CN111433355A (zh) * 2017-08-10 2020-07-17 古德T细胞有限公司 用于癌症治疗的t细胞的活化方法
CN111630602A (zh) * 2017-11-22 2020-09-04 磨石肿瘤生物技术公司 减少新抗原的接合表位呈递
US10769493B2 (en) * 2016-07-27 2020-09-08 Beijing Kuangshi Technology Co., Ltd. Method and apparatus for neural network training and construction and method and apparatus for object detection
US20200311520A1 (en) * 2019-03-29 2020-10-01 International Business Machines Corporation Training machine learning model
US10847252B2 (en) 2015-12-16 2020-11-24 Gritstone Oncology, Inc. Neoantigen identification, manufacture, and use
US10936949B2 (en) * 2017-02-24 2021-03-02 Deepmind Technologies Limited Training machine learning models using task selection policies to increase learning progress
CN112639136A (zh) * 2018-08-30 2021-04-09 蒙特利尔大学 用于鉴定肿瘤特异性抗原的基于蛋白质基因组学的方法
KR20210089094A (ko) * 2020-01-07 2021-07-15 한국과학기술원 신생항원을 스크리닝하는 방법, 시스템 및 그의 용도
CN113316818A (zh) * 2019-03-15 2021-08-27 痕准生物科技有限公司 新生抗原的鉴定方法
US11183272B2 (en) * 2018-12-21 2021-11-23 Biontech Us Inc. Method and systems for prediction of HLA class II-specific epitopes and characterization of CD4+ T cells
WO2021236854A1 (en) * 2020-05-19 2021-11-25 Gritstone Bio, Inc. Sars-cov-2 vaccines
CN113762416A (zh) * 2021-10-15 2021-12-07 南京澄实生物科技有限公司 基于多模态深度编码的抗原免疫原性预测方法和系统
US11264117B2 (en) 2017-10-10 2022-03-01 Gritstone Bio, Inc. Neoantigen identification using hotspots
US11315016B2 (en) * 2017-10-16 2022-04-26 Illumina, Inc. Deep convolutional neural networks for variant classification
US11452768B2 (en) 2013-12-20 2022-09-27 The Broad Institute, Inc. Combination therapy with neoantigen vaccine
US11514289B1 (en) * 2016-03-09 2022-11-29 Freenome Holdings, Inc. Generating machine learning models using genetic data
US20230054552A1 (en) * 2020-04-27 2023-02-23 Flagship Pioneering Innovations Vi, Llc Optimizing Proteins Using Model Based Optimizations
US11634767B2 (en) 2018-05-31 2023-04-25 Personalis, Inc. Compositions, methods and systems for processing or analyzing multi-species nucleic acid samples
US11643685B2 (en) 2016-05-27 2023-05-09 Personalis, Inc. Methods and systems for genetic analysis
US11650211B2 (en) * 2017-02-12 2023-05-16 Biontech Us Inc. HLA-based methods and compositions and uses thereof
US11725237B2 (en) 2013-12-05 2023-08-15 The Broad Institute Inc. Polymorphic gene typing and somatic change detection using sequencing data
US11814750B2 (en) 2018-05-31 2023-11-14 Personalis, Inc. Compositions, methods and systems for processing or analyzing multi-species nucleic acid samples
US11834718B2 (en) 2013-11-25 2023-12-05 The Broad Institute, Inc. Compositions and methods for diagnosing, evaluating and treating cancer by means of the DNA methylation status
US11861491B2 (en) 2017-10-16 2024-01-02 Illumina, Inc. Deep learning-based pathogenicity classifier for promoter single nucleotide variants (pSNVs)
US11927591B2 (en) * 2016-08-02 2024-03-12 Georgetown University Methods of identifying novel proteins and antigens in cancer cells
US11939637B2 (en) 2014-12-19 2024-03-26 Massachusetts Institute Of Technology Molecular biomarkers for cancer immunotherapy
US11976326B2 (en) 2013-01-17 2024-05-07 Personalis, Inc. Methods and systems for genetic analysis

Families Citing this family (60)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2014180490A1 (en) 2013-05-10 2014-11-13 Biontech Ag Predicting immunogenicity of t cell epitopes
US11338026B2 (en) 2014-09-10 2022-05-24 The University Of Connecticut Identification of immunologically protective neo-epitopes for the treatment of cancers
WO2016128060A1 (en) 2015-02-12 2016-08-18 Biontech Ag Predicting t cell epitopes useful for vaccination
AU2017367696A1 (en) 2016-12-01 2019-06-20 Nant Holdings Ip, Llc Tumor antigenicity processing and presentation
WO2018136664A1 (en) * 2017-01-18 2018-07-26 Ichan School Of Medicine At Mount Sinai Neoantigens and uses thereof for treating cancer
KR20190140935A (ko) * 2017-04-19 2019-12-20 그릿스톤 온콜로지, 인코포레이티드 신생항원 동정, 제조, 및 용도
CA3062591A1 (en) 2017-05-08 2018-11-15 Gritstone Oncology, Inc. Alphavirus neoantigen vectors
WO2019006022A1 (en) * 2017-06-27 2019-01-03 The Broad Institute, Inc. SYSTEMS AND METHODS FOR PREDICTING MHC CLASS II EPITOPE
US11634773B2 (en) 2017-07-14 2023-04-25 The Francis Crick Institute Limited Analysis of HLA alleles in tumours and the uses thereof
JP2020529470A (ja) * 2017-07-25 2020-10-08 カリフォルニア インスティチュート オブ テクノロジー トロゴサイトーシス媒介エピトープ発見
JP6928948B2 (ja) * 2017-08-09 2021-09-01 学校法人日本大学 口臭判定装置及びプログラム
JP2020532323A (ja) * 2017-09-05 2020-11-12 グリットストーン オンコロジー インコーポレイテッド Ｔ細胞療法用の新生抗原の特定法
CN109682978B (zh) * 2017-11-30 2020-07-03 四川康德赛医疗科技有限公司 一种肿瘤突变肽mhc亲和力预测方法及其应用
JP2021509823A (ja) 2018-01-04 2021-04-08 アイコニック セラピューティクス インコーポレイテッド 抗組織因子抗体、抗体薬物コンジュゲート、及び関連する方法
KR102212663B1 (ko) * 2018-05-22 2021-02-05 주식회사 석영시스템즈 목표 온도를 기반으로 하는 빌딩의 열·공조 시스템에 대한 공급 전력 제어 방법 및 장치
EP3796927A4 (en) * 2018-05-23 2022-04-20 Gritstone bio, Inc. SHARED ANTIGENS
WO2020022897A1 (en) * 2018-07-26 2020-01-30 Frame Pharmaceuticals B.V. Method of preparing subject-specific immunogenic compositions based on a neo open-reading-frame peptide database
US20210382068A1 (en) * 2018-10-02 2021-12-09 Dana-Farber Cancer Institute, Inc. Hla single allele lines
US20220088174A1 (en) * 2018-10-26 2022-03-24 Dana-Farber Cancer Institute, Inc. Genomic variants in ig gene regions and uses of same
WO2020092382A1 (en) * 2018-10-29 2020-05-07 Health Research, Inc. Cancer specific immunotherapeutic targets generated by chemotherapeutic drug treatment
JP7239300B2 (ja) * 2018-11-01 2023-03-14 浜松ホトニクス株式会社 スペクトル分析装置およびスペクトル分析方法
JP7477888B2 (ja) * 2018-11-15 2024-05-02 ノイスコム アーゲー 個別化された癌ワクチンの作製のための癌変異の選択
CA3115017C (en) 2018-11-21 2023-10-31 Nec Corporation Method and system of targeting epitopes for neoantigen-based immunotherapy
US20220076783A1 (en) * 2018-12-20 2022-03-10 Merck Sharp & Dohme Corp. Methods and Systems for the Precise Identification of Immunogenic Tumor Neoantigens
EP3897691A4 (en) * 2018-12-21 2022-08-31 The Regents of the University of California IL-10 VACCINES AND THEIR USES
EP3906045A1 (en) 2019-01-03 2021-11-10 Evaxion Biotech ApS Vaccines targeting neoepitopes
CN110675913B (zh) * 2019-01-16 2022-04-12 倍而达药业（苏州）有限公司 一种基于hla分型与结构的肿瘤新抗原的筛选方法
CN109801678B (zh) * 2019-01-25 2023-07-25 上海鲸舟基因科技有限公司 基于全转录组的肿瘤抗原预测方法及其应用
US20220098578A1 (en) * 2019-01-31 2022-03-31 Bar Ilan University Neoantigens created by aberrant-induced splicing and uses thereof in enhancing immunotherapy
EP3924971A1 (en) * 2019-02-11 2021-12-22 Flagship Pioneering Innovations VI, LLC Machine learning guided polypeptide analysis
WO2020172472A1 (en) 2019-02-20 2020-08-27 Rubius Therapeutics, Inc. Engineered erythroid cells including loadable antigen-presenting polypeptides and methods of use
CN111621564B (zh) * 2019-02-28 2022-03-25 武汉大学 一种鉴定有效肿瘤新抗原的方法
CA3132041A1 (en) * 2019-03-06 2020-09-10 Gritstone Bio, Inc. Identification of neoantigens with mhc class ii model
AU2020234003A1 (en) 2019-03-11 2021-11-11 Evaxion Biotech A/S Nucleic acid vaccination using neo-epitope encoding constructs
KR20220016137A (ko) 2019-05-30 2022-02-08 그릿스톤 바이오, 인코포레이티드 변형된 아데노바이러스
CN110172513A (zh) * 2019-05-30 2019-08-27 浙江自贸区锐赛生物医药科技有限公司 突变型免疫多肽临床入组的筛选方法
AU2020298552A1 (en) * 2019-07-02 2022-03-03 Gritstone Bio, Inc. HIV antigens and MHC complexes
WO2021011544A1 (en) 2019-07-16 2021-01-21 Gilead Sciences, Inc. Hiv vaccines and methods of making and using
EP4028763A1 (en) 2019-09-13 2022-07-20 Evaxion Biotech A/S Method for identifying t-cell epitopes
CN110706747B (zh) * 2019-09-17 2021-09-07 北京橡鑫生物科技有限公司 检测肿瘤新生抗原多肽的方法和装置
CN110456079B (zh) * 2019-09-20 2020-06-02 四川大学华西医院 Tapbp自身抗体检测试剂在制备肺癌筛查试剂盒中的用途
US20210104294A1 (en) * 2019-10-02 2021-04-08 The General Hospital Corporation Method for predicting hla-binding peptides using protein structural features
US20230072079A1 (en) 2019-12-18 2023-03-09 Evaxion Biotech A/S Nucleic acid vaccination using neo-epitope encoding constructs
CN111303268B (zh) * 2020-03-10 2021-08-27 江苏兰恩生物治疗技术研究院有限公司 子宫肌瘤新抗原、其应用和子宫肌瘤疫苗
CN115698051A (zh) 2020-04-07 2023-02-03 伊沃逊生物科技股份公司 用apc靶向单元进行新表位免疫治疗
US11920202B2 (en) 2020-04-09 2024-03-05 University Of Connecticut Unbiased identification of tumor rejection mediating neoepitopes
AU2021307553A1 (en) 2020-07-14 2023-02-16 Evaxion Biotech A/S APC targeting units for immunotherapy
EP4182928A1 (en) 2020-07-14 2023-05-24 Myneo Nv Method, system and computer program product for determining presentation likelihoods of neoantigens
JP2023541108A (ja) 2020-08-06 2023-09-28 グリットストーン バイオ インコーポレイテッド マルチエピトープワクチンカセット
US20220139498A1 (en) * 2020-11-04 2022-05-05 Basf Corporation Apparatuses, systems, and methods for extracting meaning from dna sequence data using natural language processing (nlp)
US20230197192A1 (en) * 2020-11-06 2023-06-22 Amazon Technologies, Inc. Selecting neoantigens for personalized cancer vaccine
US20220319635A1 (en) * 2021-04-05 2022-10-06 Nec Laboratories America, Inc. Generating minority-class examples for training data
CN113129998B (zh) * 2021-04-23 2022-06-21 云测智能科技有限公司 临床个体化肿瘤新抗原的预测模型构建方法
JP2024518302A (ja) 2021-04-29 2024-05-01 イェダ リサーチ アンド デベロップメント カンパニー リミテッド Ｒａｓ誘導性再発性ネオアンチゲンに対するｔ細胞受容体およびそれらを同定するための方法
BR112023022485A2 (pt) 2021-04-30 2024-01-09 Centre Hospitalier Univ Vaudois Chuv Expansão de linfócitos em um único recipiente
CN113807468B (zh) * 2021-10-15 2022-05-27 南京澄实生物科技有限公司 基于多模态深度编码的hla抗原呈递预测方法和系统
WO2023177655A1 (en) 2022-03-14 2023-09-21 Generation Bio Co. Heterologous prime boost vaccine compositions and methods of use
TWI805290B (zh) * 2022-03-28 2023-06-11 臺北醫學大學 用於預測肺腺癌是否具有表皮生長因子受體突變的方法
US20230368872A1 (en) * 2022-05-16 2023-11-16 Spencer Health Solutions, Llc Alert decision support system based on patient quality of life survey information and related methods and computer program products
WO2024097864A1 (en) 2022-11-02 2024-05-10 Tigen Pharma Sa Expansion of lymphocytes

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020155447A1 (en) * 2000-04-04 2002-10-24 Maurice Zauderer Gene differentially expressed in breast cancer, and encoded polypeptides
US20030171290A1 (en) * 2000-08-03 2003-09-11 Carr Francis J. Peptides presented by cells
US20150278441A1 (en) * 2014-03-25 2015-10-01 Nec Laboratories America, Inc. High-order semi-Restricted Boltzmann Machines and Deep Models for accurate peptide-MHC binding prediction
US20160347798A1 (en) * 2014-01-27 2016-12-01 Molecular Templates, Inc. Mhc class i epitope delivering polypeptides and cell-targeted molecules for direct cell killing and immune stimulation via mhc class i presentation and methods regarding the same
US20170116371A1 (en) * 2015-10-23 2017-04-27 Google Inc. Neural network for processing aptamer data
US20170221176A1 (en) * 2016-01-29 2017-08-03 Fotonation Limited Convolutional neural network

Family Cites Families (230)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4235871A (en) 1978-02-24 1980-11-25 Papahadjopoulos Demetrios P Method of encapsulating biologically active materials in lipid vesicles
US4722848A (en) 1982-12-08 1988-02-02 Health Research, Incorporated Method for immunizing animals with synthetically modified vaccinia virus
US4452773A (en) 1982-04-05 1984-06-05 Canadian Patents And Development Limited Magnetic iron-dextran microspheres
US4501728A (en) 1983-01-06 1985-02-26 Technology Unlimited, Inc. Masking of liposomes from RES recognition
GB8311018D0 (en) 1983-04-22 1983-05-25 Amersham Int Plc Detecting mutations in dna
US5019369A (en) 1984-10-22 1991-05-28 Vestar, Inc. Method of targeting tumors in humans
US4795698A (en) 1985-10-04 1989-01-03 Immunicon Corporation Magnetic-polymer particles
US4837028A (en) 1986-12-24 1989-06-06 Liposome Technology, Inc. Liposomes with enhanced circulation time
US5204244A (en) 1987-10-27 1993-04-20 Oncogen Production of chimeric antibodies by homologous recombination
AU4746590A (en) 1988-12-28 1990-08-01 Stefan Miltenyi Methods and materials for high gradient magnetic separation of biological materials
US5703055A (en) 1989-03-21 1997-12-30 Wisconsin Alumni Research Foundation Generation of antibodies through lipid mediated DNA delivery
US5130538A (en) 1989-05-19 1992-07-14 John B. Fenn Method of producing multiply charged ions and for determining molecular weights of molecules by use of the multiply charged ions of molecules
US5581080A (en) 1989-05-19 1996-12-03 Fenn; John B. Method for determining molecular weight using multiply charged ions
AU5815590A (en) 1989-05-19 1990-12-18 John B. Fenn Multiply charged ions and a method for determining the molecular weight of large molecules
FR2650840B1 (fr) 1989-08-11 1991-11-29 Bertin & Cie Procede rapide de detection et/ou d'identification d'une seule base sur une sequence d'acide nucleique, et ses applications
EP0800830A3 (en) 1989-11-03 1999-03-17 Vanderbilt University Method of in vivo delivery of functioning foreign genes
CA2076386C (en) 1990-02-26 2003-04-22 David S. Hogness Identification and expression of insect steroid receptor dna sequence
US5279833A (en) 1990-04-04 1994-01-18 Yale University Liposomal transfection of nucleic acids into animal cells
US5204253A (en) 1990-05-29 1993-04-20 E. I. Du Pont De Nemours And Company Method and apparatus for introducing biological substances into living cells
US5200084A (en) 1990-09-26 1993-04-06 Immunicon Corporation Apparatus and methods for magnetic separation
US6004744A (en) 1991-03-05 1999-12-21 Molecular Tool, Inc. Method for determining nucleotide identity through extension of immobilized primer
AU4528493A (en) 1992-06-04 1994-01-04 Regents Of The University Of California, The In vivo gene therapy with intron-free sequence of interest
DE4444229C2 (de) 1994-03-10 1996-07-25 Bruker Franzen Analytik Gmbh Verfahren und Vorrichtungen zur Elektrosprüh-Ionisierung für speichernde Massenspektometer
US5608217A (en) 1994-03-10 1997-03-04 Bruker-Franzen Analytik Gmbh Electrospraying method for mass spectrometric analysis
US5534615A (en) 1994-04-25 1996-07-09 Genentech, Inc. Cardiac hypertrophy factor and uses therefor
US5827642A (en) 1994-08-31 1998-10-27 Fred Hutchinson Cancer Research Center Rapid expansion method ("REM") for in vitro propagation of T lymphocytes
US8114414B2 (en) 1994-11-08 2012-02-14 The Trustees Of The University Of Pennsylvania Compositions and methods for treatment of cervical cancer
US8956621B2 (en) 1994-11-08 2015-02-17 The Trustees Of The University Of Pennsylvania Compositions and methods for treatment of cervical dysplasia
US6071890A (en) 1994-12-09 2000-06-06 Genzyme Corporation Organ-specific targeting of cationic amphiphile/DNA complexes for gene therapy
US5849589A (en) 1996-03-11 1998-12-15 Duke University Culturing monocytes with IL-4, TNF-α and GM-CSF TO induce differentiation to dendric cells
JP2001520013A (ja) 1997-10-10 2001-10-30 ビーエーエスエフ アクチェンゲゼルシャフト Ｔ細胞受容体−会合分子（ｔａｒｍｓ）及びその使用方法
DE19937828C1 (de) 1999-08-11 2000-10-05 Smb Schwede Maschinenbau Gmbh Schweißkopf für eine Bandumreifungsmaschine
AU7508100A (en) 1999-10-04 2001-05-10 Jens Andersen Human seizure related proteins
CA2404253C (en) 2000-03-22 2014-05-13 Rohm And Haas Company Novel ecdysone receptor-based inducible gene expression system
US9012141B2 (en) 2000-03-27 2015-04-21 Advaxis, Inc. Compositions and methods comprising KLK3 of FOLH1 antigen
AU2001270941A1 (en) 2000-06-09 2001-12-17 Mds Proteomics, Inc. Labeling of proteomic samples during proteolysis for quantitation and sample multiplexing
AU2002234171A1 (en) 2000-10-25 2002-05-15 Mds Proteomics, Inc. Detection of modified amino acids by mass spectrometry
US7981420B2 (en) 2000-12-22 2011-07-19 Max-Planck-Gesellschaft Zur Foederung Der Wissenschaften E.V. Therapeutic use of antibodies directed against repulsive guidance molecule (RGM)
US8771702B2 (en) 2001-03-26 2014-07-08 The Trustees Of The University Of Pennsylvania Non-hemolytic LLO fusion proteins and methods of utilizing same
AU2002256173A1 (en) 2001-04-09 2002-10-21 Mds Proteomics, Inc. Methods and systems for searching genomic databases
US7731648B2 (en) 2001-07-25 2010-06-08 Aduro Biotech Magnetic nanoscale particle compositions, and therapeutic methods related thereto
WO2003038055A2 (en) 2001-10-31 2003-05-08 Mds Proteomics, Inc. Proteins involved in regulation of adipocytes and uses related thereto
DE10211088A1 (de) 2002-03-13 2003-09-25 Ugur Sahin Differentiell in Tumoren exprimierte Genprodukte und deren Verwendung
AU2003232485A1 (en) 2002-04-18 2003-10-27 Mtm Laboratories Ag Neopeptides and methods useful for detection and treatment of cancer
US20050221350A1 (en) 2002-05-29 2005-10-06 Toni Weinschenk Method for identifying immunoreactive peptides
US7176022B2 (en) 2002-12-20 2007-02-13 Cell Genesys, Inc. Directly injectable formulations which provide enhanced cryoprotection of cell products
MXPA05008340A (es) 2003-02-06 2006-03-13 Cerus Corp Microbios de vida libre modificados, composiciones de vacuna y metodos de uso de los mismos.
KR101192652B1 (ko) 2003-02-06 2012-10-19 앤저 테라퓨틱스 인코퍼레이티드 비포식 세포로의 침투가 약화된 리스테리아, 리스테리아를 포함하는 백신, 및 그것의 사용방법
US7695725B2 (en) 2003-02-06 2010-04-13 Aduro Biotech Modified free-living microbes, vaccine compositions and methods of use thereof
CN1780850A (zh) 2003-02-28 2006-05-31 抗基因公司 凝集素在促进糖蛋白和抗原分子的低聚反应中的用途
US20040197312A1 (en) 2003-04-02 2004-10-07 Marina Moskalenko Cytokine-expressing cellular vaccine combinations
DE10341812A1 (de) 2003-09-10 2005-04-07 Ganymed Pharmaceuticals Ag Differentiell in Tumoren exprimierte Genprodukte und deren Verwendung
DE10344799A1 (de) 2003-09-26 2005-04-14 Ganymed Pharmaceuticals Ag Identifizierung von Oberflächen-assoziierten Antigenen für die Tumordiagnose und -therapie
MXPA06007144A (es) 2003-12-24 2007-01-31 Cerus Corp Moleculas de acido nucleico recombinante que codifican proteinas de fusion que comprenden polipeptidos de senal secretoria bacterial y de antigenos, casetes de expresion y bacterias, y metodos para usar las mismas.
US7842289B2 (en) 2003-12-24 2010-11-30 Aduro Biotech Recombinant nucleic acid molecules, expression cassettes, and bacteria, and methods of use thereof
DE102004023187A1 (de) 2004-05-11 2005-12-01 Ganymed Pharmaceuticals Ag Identifizierung von Oberflächen-assoziierten Antigenen für die Tumordiagnose und -therapie
US20070265818A1 (en) 2004-08-24 2007-11-15 Searete Llc, A Limited Liability Corporation Of The State Of Delaware Computational methods and systems for heightening cell-mediated immune response
US20060095241A1 (en) 2004-10-29 2006-05-04 Microsoft Corporation Systems and methods that utilize machine learning algorithms to facilitate assembly of aids vaccine cocktails
US7220549B2 (en) 2004-12-30 2007-05-22 Helicos Biosciences Corporation Stabilizing a nucleic acid for nucleic acid sequencing
US7283337B2 (en) 2005-03-04 2007-10-16 Headway Technologies, Inc. Abutted exchange bias design for sensor stabilization
DE102005013846A1 (de) 2005-03-24 2006-10-05 Ganymed Pharmaceuticals Ag Identifizierung von Oberflächen-assoziierten Antigenen für die Tumordiagnose und -therapie
PT1806358E (pt) 2005-09-05 2010-05-28 Immatics Biotechnologies Gmbh Peptídeos associados a tumor ligando promiscuamente às moléculas do antigénio de leucócitos humanos (hla) da classe ii
EP1762575A1 (en) 2005-09-12 2007-03-14 Ganymed Pharmaceuticals AG Identification of tumor-associated antigens for diagnosis and therapy
US9732131B2 (en) 2006-02-27 2017-08-15 Calviri, Inc. Identification and use of novopeptides for the treatment of cancer
AU2007220042A1 (en) 2006-02-27 2007-09-07 Arizona Board Of Regents For And On Behalf Of Arizona State University Identification and use of novopeptides for the treatment of cancer
US7935804B2 (en) 2006-03-01 2011-05-03 Aduro Biotech Engineered Listeria and methods of use thereof
ES2532942T3 (es) 2006-03-01 2015-04-06 Aduro Biotech Listeria obtenida por ingeniería genética y métodos de uso de la misma
US7919079B2 (en) 2006-03-31 2011-04-05 Biosante Pharmaceuticals, Inc. Cancer immunotherapy compositions and methods of use
US8768629B2 (en) 2009-02-11 2014-07-01 Caris Mpi, Inc. Molecular profiling of tumors
DE102006032362A1 (de) 2006-07-13 2008-01-17 Khd Humboldt Wedag Gmbh Rollenpresse insbesondere zur Gutbettzerkleinerung
US8926993B2 (en) 2006-07-17 2015-01-06 Aduro Biotech Methods and compositions using Listeria for enhancing immunogenicity by prime boost
US8121797B2 (en) 2007-01-12 2012-02-21 Microsoft Corporation T-cell epitope prediction
US9085638B2 (en) 2007-03-07 2015-07-21 The Johns Hopkins University DNA vaccine enhancement with MHC class II activators
US20090004213A1 (en) 2007-03-26 2009-01-01 Immatics Biotechnologies Gmbh Combination therapy using active immunotherapy
SI2567707T1 (sl) 2007-07-27 2017-01-31 Immatics Biotechnologies Gmbh Sestava s tumorjem povezanih peptidov in zadevnih cepiv proti raku
CA2694805C (en) 2007-07-27 2014-09-09 Immatics Biotechnologies Gmbh Immunogenic epitopes of tumour-associated antigens
PT2660248E (pt) 2007-07-27 2015-10-12 Immatics Biotechnologies Gmbh Nova imunoterapia para tumores neuronais e cerebrais
US20100286070A1 (en) 2007-09-14 2010-11-11 Gert Verheyden Affinity tag
EP2060583A1 (en) 2007-10-23 2009-05-20 Ganymed Pharmaceuticals AG Identification of tumor-associated markers for diagnosis and therapy
ES2640216T3 (es) 2007-12-07 2017-11-02 Miltenyi Biotec Gmbh Sistemas y métodos para procesamiento de células
EP2091046A1 (en) 2008-02-15 2009-08-19 Thomson Licensing Presentation system and method for controlling the same
PT2113253E (pt) 2008-04-30 2010-06-15 Immatics Biotechnologies Gmbh Formulações novas de peptídeos associados a tumores que se ligam a moléculas de classe i ou ii do antígeno leucocitário humano (hla) para vacinas
US20120164718A1 (en) 2008-05-06 2012-06-28 Innovative Micro Technology Removable/disposable apparatus for MEMS particle sorting device
NO2119726T3 (ja) 2008-05-14 2015-05-23
US20150366955A9 (en) 2009-11-11 2015-12-24 Advaxis, Inc. Compositions and methods for prevention of escape mutation in the treatment of her2/neu over-expressing tumors
US9650639B2 (en) 2008-05-19 2017-05-16 Advaxis, Inc. Dual delivery system for heterologous antigens
US9017660B2 (en) 2009-11-11 2015-04-28 Advaxis, Inc. Compositions and methods for prevention of escape mutation in the treatment of Her2/neu over-expressing tumors
US20110129499A1 (en) 2008-05-19 2011-06-02 Paulo Maciag Dual delivery system for heterologous antigens
CA2724649A1 (en) 2008-05-19 2009-11-26 Anza Therapeutics, Inc. Compositions comprising prfa* mutant listeria and methods of use thereof
US9084747B2 (en) 2009-11-11 2015-07-21 Advaxis, Inc. Compositions and methods for prevention of escape mutation in the treatment of HER2/NEU over-expressing tumors
US20140234370A1 (en) 2009-11-11 2014-08-21 Advaxis, Inc. Compositions and methods for prevention of escape mutation in the treatment of her2/neu over-expressing tumors
US8840881B2 (en) 2008-08-28 2014-09-23 Aduro Gvax Inc. Methods and compositions for treating prostate cancer or inducing a humoral immune response against prostate cancer
EP3216874A1 (en) 2008-09-05 2017-09-13 TOMA Biosciences, Inc. Methods for stratifying and annotating cancer drug treatment options
PT2172211E (pt) 2008-10-01 2015-03-09 Immatics Biotechnologies Gmbh Composição de peptídeos associados a tumores e vacina anti -cancro relacionada para o tratamento de glioblastoma (gbm) e outros tipos de câncer
HUE045093T2 (hu) 2008-11-21 2019-12-30 Kobenhavns Univ University Of Copenhagen Immunválasz kiváltása
SI2403528T1 (sl) 2009-03-02 2016-07-29 Aduro Biotech Holdings, Europe B.V. Protitelesa proti proliferacijo inducirajočemu ligandu (april)
ES2622228T3 (es) 2009-03-10 2017-07-06 Baylor Research Institute Vacunas antivíricas dirigidas a células presentadoras de antígenos
WO2010106535A1 (en) 2009-03-15 2010-09-23 Technion Research And Development Foundation Ltd. Soluble hla complexes for use in disease diagnosis
WO2010115118A2 (en) 2009-04-03 2010-10-07 Antigenics, Inc. Methods for preparing and using multichaperone-antigen complexes
EP2309262A1 (en) 2009-10-06 2011-04-13 Max-Planck-Gesellschaft zur Förderung der Wissenschaften e.V. Method for quantifying biomolecules
US20110223187A1 (en) 2010-02-15 2011-09-15 Vafa Shahabi Live listeria-based vaccines for central nervous system therapy
GB201004551D0 (en) 2010-03-19 2010-05-05 Immatics Biotechnologies Gmbh NOvel immunotherapy against several tumors including gastrointestinal and gastric cancer
GB201004575D0 (en) 2010-03-19 2010-05-05 Immatics Biotechnologies Gmbh Composition of tumor associated peptides and related anti cancer vaccine for the treatment of gastric cancer and other cancers
CA2797868C (en) 2010-05-14 2023-06-20 The General Hospital Corporation Compositions and methods of identifying tumor specific neoantigens
US9161974B2 (en) 2010-05-23 2015-10-20 Aduro Biotech, Inc. Methods and compositions using listeria for adjuvant treatment of cancer
CA2800535A1 (en) 2010-05-24 2011-12-01 Phosimmune, Inc. Class i mhc phosphopeptides for cancer immunotherapy and diagnosis
EP2591001B1 (en) 2010-07-09 2021-11-17 Aduro Biotech Holdings, Europe B.V. Agonistic antibody to cd27
EP2616482A1 (en) 2010-09-14 2013-07-24 Max-Planck-Gesellschaft zur Förderung der Wissenschaften e.V. Hot1 and uses thereof
GB201015765D0 (en) 2010-09-21 2010-10-27 Immatics Biotechnologies Gmbh Use of myeloid cell biomarkers for the diagnosis of cancer
WO2012138377A2 (en) 2010-10-01 2012-10-11 Trustees Of The University Of Pennsylvania The use of listeria vaccine vectors to reverse vaccine unresponsiveness in parasitically infected individuals
DK2640842T3 (en) 2010-11-17 2018-08-13 Aduro Biotech Inc Methods and compositions for inducing an immune response to EGFRVIII
GB201021289D0 (en) 2010-12-15 2011-01-26 Immatics Biotechnologies Gmbh Novel biomarkers for a prediction of the outcome of an immunotherapy against cancer
KR20210131432A (ko) 2010-12-30 2021-11-02 파운데이션 메디신 인코포레이티드 종양 샘플의 다유전자 분석의 최적화
CA2829960A1 (en) 2011-03-11 2012-09-20 John Rothman Listeria-based adjuvants
EP2508537A1 (en) 2011-04-04 2012-10-10 Max-Planck-Gesellschaft zur Förderung der Wissenschaften e.V. Quantitative standard for mass spectrometry of proteins
AU2012261237B2 (en) 2011-05-24 2017-06-01 BioNTech SE Individualized vaccines for cancer
WO2013138337A1 (en) 2012-03-12 2013-09-19 Advaxis Suppressor cell function inhibition following listeria vaccine treatment
WO2013158611A1 (en) 2012-04-16 2013-10-24 Agenus Inc. Methods and compositions for the treatment of glioblastomas
CA2879024A1 (en) 2012-07-12 2014-01-16 Persimmune, Inc. Personalized cancer vaccines and adoptive immune cell therapies
EP2893061B1 (en) 2012-09-05 2019-05-22 Arizona Board of Regents, a Body Corporate of the State of Arizona acting for and on behalf of Arizona State University Methods for discovering therapeutic targets
JP6499079B2 (ja) 2012-11-13 2019-04-10 バイオエヌテック アーゲーＢｉｏＮＴｅｃｈ ＡＧ クローディンを発現するガン疾患を処置するための剤
EP2925348B1 (en) 2012-11-28 2019-03-06 BioNTech RNA Pharmaceuticals GmbH Individualized vaccines for cancer
AU2013358892B2 (en) 2012-12-13 2018-06-21 Aduro Biotech, Inc. Compositions comprising cyclic purine dinucleotides having defined stereochemistries and methods for their preparation and use
ES2939121T3 (es) 2012-12-19 2023-04-19 Max Planck Gesellschaft Recipiente de reacción para preparación de muestras
EP2938627B1 (en) 2012-12-27 2019-03-20 Aduro Biotech, Inc. Signal peptide fusion partners facilitating listerial expression of antigenic sequences and methods of preparation and use thereof
JP6450690B2 (ja) 2013-02-15 2019-01-09 ザ リージェンツ オブ ザ ユニバーシティ オブ カリフォルニア キメラ抗原受容体及びその使用方法
AU2014251207B2 (en) 2013-04-07 2019-06-13 Dana-Farber Cancer Institute, Inc. Compositions and methods for personalized neoplasia vaccines
WO2014180490A1 (en) 2013-05-10 2014-11-13 Biontech Ag Predicting immunogenicity of t cell epitopes
AU2014268836B2 (en) 2013-05-18 2018-08-02 Aduro Biotech, Inc. Compositions and methods for activating "stimulator of interferon gene"-dependent signalling
EP3008204B1 (en) * 2013-06-10 2024-03-13 Iogenetics, LLC. Mathematical processes for determination of peptidase cleavage
EP3777882A1 (en) 2013-07-30 2021-02-17 BioNTech SE Tumor antigens for determining cancer therapy
CA2919567A1 (en) 2013-07-30 2015-02-05 Biontech Ag Tumor antigens for determining cancer therapy
AU2014296887A1 (en) 2013-08-02 2016-01-28 Aduro Biotech Holdings, Europe B.V. Combining CD27 agonists and immune checkpoint inhibition for immune stimulation
TWI777194B (zh) 2013-08-05 2022-09-11 德商伊瑪提克斯生物科技有限公司 新穎肽類，細胞及其用於治療多種腫瘤的用途，其製造方法及包含其等之醫藥組成物（一）
EP3415639B8 (en) 2013-08-19 2023-07-12 Shuwen Biotech Co., Ltd. Methods and kits for the molecular subtyping of tumors
SG11201601060RA (en) 2013-08-19 2016-03-30 Biontech Diagnostics Gmbh Methods and kits for the molecular subtyping of tumors
WO2015030585A2 (en) 2013-08-27 2015-03-05 Academisch Ziekenhuis Leiden H.O.D.N. Lumc Methods for detecting post-translationally modified lysines in a polypeptide
AU2014317009A1 (en) 2013-09-05 2016-03-10 Aduro Biotech Holdings, Europe B.V. CD70-binding peptides and method, process and use relating thereto
NL2011406C2 (en) 2013-09-06 2015-03-10 Bionovion Holding B V Method for obtaining april-binding peptides, process for producing the peptides, april-binding peptides obtainable with said method/process and use of the april-binding peptides.
JP6363179B2 (ja) 2013-09-26 2018-07-25 バイオンテック アーゲー Ｒｎａによるシェルを含む粒子
EP3060679B1 (en) 2013-10-25 2019-06-12 BioNTech Diagnostics GmbH Method and kit for determining whether a subject shows an immune response
WO2015058780A1 (en) 2013-10-25 2015-04-30 Biontech Ag Method and kit for determining whether a subject shows an immune response
KR102006527B1 (ko) 2013-11-01 2019-08-02 화이자 인코포레이티드 전립선-연관 항원의 발현을 위한 벡터
GB201319446D0 (en) 2013-11-04 2013-12-18 Immatics Biotechnologies Gmbh Personalized immunotherapy against several neuronal and brain tumors
US11725237B2 (en) 2013-12-05 2023-08-15 The Broad Institute Inc. Polymorphic gene typing and somatic change detection using sequencing data
CN106456724A (zh) 2013-12-20 2017-02-22 博德研究所 使用新抗原疫苗的联合疗法
CA2935214A1 (en) 2014-01-02 2015-07-09 Memorial Sloan Kettering Cancer Center Determinants of cancer response to immunotherapy
EP3099708B1 (en) 2014-01-29 2019-09-04 University Health Network Methods and compositions for producing a cell expressing a t cell receptor
ES2699753T3 (es) 2014-01-29 2019-02-12 Biontech Ag Mimótopos peptídicos de claudina 18.2 y sus usos
CN106456726A (zh) 2014-02-18 2017-02-22 阿德瓦希斯公司 生物标志物导向的多靶点免疫治疗
MX2016011114A (es) 2014-02-25 2017-02-20 Advaxis Inc Composiciones y metodos para el tratamiento de tumores que sobreexpresan her2/neu.
US20170080064A1 (en) 2014-03-05 2017-03-23 Advaxis, Ins. Methods and compositions for increasing a t-effector cell to regulatory t cell ratio
SG10202011841WA (en) 2014-04-24 2021-01-28 Advaxis Inc Recombinant listeria vaccine strains and methods of producing the same
GB201408255D0 (en) 2014-05-09 2014-06-25 Immatics Biotechnologies Gmbh Novel immunotherapy against several tumours of the blood, such as acute myeloid leukemia (AML)
EP2944955A1 (en) 2014-05-13 2015-11-18 Max-Planck-Gesellschaft zur Förderung der Wissenschaften e.V. Benchmark for LC-MS systems
WO2015172843A1 (en) 2014-05-16 2015-11-19 Biontech Diagnostics Gmbh Methods and kits for the diagnosis of cancer
GB201411037D0 (en) 2014-06-20 2014-08-06 Immatics Biotechnologies Gmbh Novel immunotherapy against several tumors of the blood, in particular chronic lymphoid leukemai (CLL)
US10000533B2 (en) 2014-06-20 2018-06-19 Immatics Biotechnologies Gmbh Immunotherapy against several tumors of the blood, in particular chronic lymphoid leukemia (CLL)
WO2016011357A1 (en) 2014-07-18 2016-01-21 Advaxis, Inc. Combination of a pd-1 antagonist and a listeria-based vaccine for treating prostate cancer
AU2015289449A1 (en) 2014-07-18 2017-02-09 Advaxis, Inc. Listeria-based immunogenic compositions for eliciting anti-tumor responses
WO2016011320A1 (en) 2014-07-18 2016-01-21 Advaxis, Inc. Bivalent listeria-based delivery system of heterologous antigens
CA2960834A1 (en) 2014-09-10 2016-03-17 Genentech, Inc. Immunogenic mutant peptide screening platform
EP3193892A4 (en) 2014-09-14 2018-09-12 Washington University Personalized cancer vaccines and methods therefor
US10590483B2 (en) 2014-09-15 2020-03-17 Abvitro Llc High-throughput nucleotide library sequencing
KR20210032011A (ko) 2014-09-17 2021-03-23 더 존스 홉킨스 유니버시티 항원-특이적 t 세포 식별, 농축, 및/또는 증식을 위한 시약 및 방법
EP3201322A4 (en) 2014-10-03 2018-05-30 Yale University Innate immune system modification for anticancer therapy
WO2016062323A1 (en) 2014-10-20 2016-04-28 Biontech Ag Methods and compositions for diagnosis and treatment of cancer
EP3215168B1 (en) 2014-10-31 2023-08-02 The Trustees of the University of Pennsylvania Altering gene expression in modified t cells and uses thereof
DE102014116335A1 (de) 2014-11-10 2016-05-12 Thyssenkrupp Ag Verbundwerkstoff, Verbundwerkstoffprodukt, Verfahren zu deren Herstellung und Verwendungen dafür
MA40737A (fr) 2014-11-21 2017-07-04 Memorial Sloan Kettering Cancer Center Déterminants de la réponse d'un cancer à une immunothérapie par blocage de pd-1
EP3234130B1 (en) 2014-12-19 2020-11-25 The Broad Institute, Inc. Methods for profiling the t-cell- receptor repertoire
MA41217A (fr) 2014-12-19 2017-10-24 Advaxis Inc Polythérapies ayant des souches de listeria recombinées
EP3234193B1 (en) 2014-12-19 2020-07-15 Massachusetts Institute of Technology Molecular biomarkers for cancer immunotherapy
GB201501017D0 (en) 2014-12-23 2015-03-04 Immatics Biotechnologies Gmbh Novel peptides and combination of peptides for use in immunotherapy against hepatocellular carcinoma (HCC) and other cancers
SG10202003171TA (en) 2015-01-08 2020-05-28 BioNTech SE Agonistic tnf receptor binding agents
NL2014108B1 (en) 2015-01-09 2016-09-30 Aduro Biotech Holdings Europe B V Altered april binding antibodies.
WO2016126876A2 (en) 2015-02-03 2016-08-11 Advaxis, Inc. Listeria-based adjuvants
US20160220652A1 (en) 2015-02-03 2016-08-04 Advaxis, Inc. Methods of using recombinant listeria vaccine strains in disease immunotherapy
EP3256846A1 (en) 2015-02-09 2017-12-20 Max-Planck-Gesellschaft zur Förderung der Wissenschaften e.V. Means and methods for minimizing swept and dead volumes in chromatographic applications
WO2016128060A1 (en) 2015-02-12 2016-08-18 Biontech Ag Predicting t cell epitopes useful for vaccination
WO2016131875A1 (en) 2015-02-17 2016-08-25 Biontech Diagnostics Gmbh Methods and kits for the molecular subtyping of bladder cancer
MA41644A (fr) 2015-03-03 2018-01-09 Advaxis Inc Compositions à base de listeria comprenant un système d'expression de minigènes codant pour des peptides, et leurs procédés d'utilisation
EP3267969A1 (en) 2015-03-09 2018-01-17 King's College London Combination therapy with rar alpha agonists for enhancing th1 response
WO2016145578A1 (en) 2015-03-13 2016-09-22 Syz Cell Therapy Co. Methods of cancer treatment using activated t cells
GB201504502D0 (en) 2015-03-17 2015-04-29 Immatics Biotechnologies Gmbh Novel peptides and combination of peptides for use in immunotherapy against pancreatic cancer and other cancers
WO2016154412A2 (en) 2015-03-26 2016-09-29 The Trustees Of The University Of Pennsylvania Combination of a pd-1 antagonist and a listeria based vaccine for treating pancreatic cancer
GB201505305D0 (en) 2015-03-27 2015-05-13 Immatics Biotechnologies Gmbh Novel Peptides and combination of peptides for use in immunotherapy against various tumors
GB201505585D0 (en) 2015-03-31 2015-05-13 Immatics Biotechnologies Gmbh Novel peptides and combination of peptides and scaffolds for use in immunotherapy against renal cell carinoma (RCC) and other cancers
WO2016155809A1 (en) 2015-03-31 2016-10-06 Biontech Rna Pharmaceuticals Gmbh Lipid particle formulations for delivery of rna and water-soluble therapeutically effective compounds to a target cell
US9808516B2 (en) 2015-04-13 2017-11-07 Aduro Biotech, Inc. Immunogenic fusion proteins for the treatment of cancer
JP6782253B2 (ja) 2015-04-13 2020-11-11 アデュロ バイオテック，インコーポレイテッド 上皮増殖因子受容体変異型ｉｉｉ−メソテリン融合物及びこれを使用する方法
WO2016172624A1 (en) 2015-04-22 2016-10-27 Agenus Inc. Methods for treating cancer
GB201507030D0 (en) 2015-04-24 2015-06-10 Immatics Biotechnologies Gmbh Immunotherapy against lung cancers, in particular NSCLC
ES2836273T3 (es) 2015-04-27 2021-06-24 Cancer Research Tech Ltd Método para el tratamiento del cáncer
GB201507719D0 (en) 2015-05-06 2015-06-17 Immatics Biotechnologies Gmbh Novel peptides and combination of peptides and scaffolds thereof for use in immunotherapy against colorectal carcinoma (CRC) and other cancers
EP3292152A1 (en) 2015-05-07 2018-03-14 Agenus Inc. Anti-ox40 antibodies and methods of use thereof
WO2016180467A1 (en) 2015-05-11 2016-11-17 Biontech Cell & Gene Therapies Gmbh Enhancing the effect of car-engineered t cells by means of nucleic acid vaccination
CA2984643A1 (en) 2015-05-13 2016-11-17 Agenus Inc. Vaccines for treatment and prevention of cancer
WO2016183361A1 (en) 2015-05-13 2016-11-17 Advaxis, Inc. Immunogenic listeria-based compositions comprising truncated acta-antigen fusions and methods of use thereof
TWI806815B (zh) 2015-05-20 2023-07-01 美商博德研究所有限公司 共有之ｇａｔａ３相關之腫瘤特異性新抗原
US10293637B2 (en) 2015-05-21 2019-05-21 Burley Design Llc Bicycle quick release cam lever with biased closure
KR20180026670A (ko) 2015-05-26 2018-03-13 어드박시스, 인크. 맞춤형 전달 벡터-기반의 면역 요법 및 이의 용도
JP6900323B2 (ja) 2015-05-29 2021-07-07 アジェナス インコーポレイテッド 抗ｃｔｌａ−４抗体およびその使用方法
GB201510771D0 (en) 2015-06-19 2015-08-05 Immatics Biotechnologies Gmbh Novel peptides and combination of peptides for use in immunotherapy and methods for generating scaffolds for the use against pancreatic cancer
KR20180027501A (ko) 2015-06-24 2018-03-14 어드박시스, 인크. 맞춤형 전달 벡터-기반 면역 요법을 위한 제조 장치 및 공정
GB201511191D0 (en) 2015-06-25 2015-08-12 Immatics Biotechnologies Gmbh T-cell epitopes for the immunotherapy of myeloma
GB201511546D0 (en) 2015-07-01 2015-08-12 Immatics Biotechnologies Gmbh Novel peptides and combination of peptides for use in immunotherapy against ovarian cancer and other cancers
GB201511792D0 (en) 2015-07-06 2015-08-19 Immatics Biotechnologies Gmbh Novel peptides and combination of peptides for use in immunotherapy against esopageal cancer and other cancers
EP3115369A1 (en) 2015-07-09 2017-01-11 Max-Planck-Gesellschaft zur Förderung der Wissenschaften e.V. Peptide purification using mixed-phase solid phase extraction material
MY189596A (en) 2015-07-15 2022-02-18 Immatics Biotechnologies Gmbh A novel peptides for use in immunotherapy against epithelial ovarian cancer and other cancers
WO2017017232A1 (en) 2015-07-29 2017-02-02 MAX-PLANCK-Gesellschaft zur Förderung der Wissenschaften e.V. Means and methods for a sample preparation, especially for mass spectrometry
WO2017024006A1 (en) 2015-08-03 2017-02-09 The Johns Hopkins University Personalized, allogeneic cell therapy of cancer
GB201513921D0 (en) 2015-08-05 2015-09-23 Immatics Biotechnologies Gmbh Novel peptides and combination of peptides for use in immunotherapy against prostate cancer and other cancers
WO2017030956A1 (en) 2015-08-14 2017-02-23 Agenus Inc. Method of inducing a t-cell response to phosphopeptides using nucleic acids encoding phosphopeptide mimetics
US20170136108A1 (en) 2015-08-28 2017-05-18 Immatics Biotechnologies Gmbh Novel peptides, combination of peptides and scaffolds for use in immunotherapeutic treatment of various cancers
GB201515321D0 (en) 2015-08-28 2015-10-14 Immatics Biotechnologies Gmbh Novel peptides, combination of peptides and scaffolds for use in immunotherapeutic treatment of various cancers
EP3344656A1 (en) 2015-09-01 2018-07-11 Agenus Inc. Anti-pd-1 antibodies and methods of use thereof
EP3362103A4 (en) 2015-10-12 2020-02-05 Nantomics, LLC COMPOSITIONS AND METHODS USING VIRAL CANCER NEOEPITOPES
EP3390434A2 (en) 2015-12-14 2018-10-24 Bellicum Pharmaceuticals, Inc. Dual controls for therapeutic cell activation or elimination
TWI765875B (zh) 2015-12-16 2022-06-01 美商磨石生物公司 新抗原辨識、製造及用途
WO2017184590A1 (en) 2016-04-18 2017-10-26 The Broad Institute Inc. Improved hla epitope prediction
BR112019010565A2 (pt) 2016-11-23 2019-09-17 Gritstone Oncology Inc aplicação viral de neoantígenos
KR20180063481A (ko) 2016-12-02 2018-06-12 주식회사 원소프트다임 체지방을 측정하는 클라우드 기반 휴대용 장치를 이용하여 헬스케어 정보를 제공하는 방법 및 이를 이용하는 장치
CN110799196B (zh) * 2017-03-31 2024-02-13 行动基因(智财)有限公司 致免疫性的癌症特异抗原决定位的排名系统
KR20190140935A (ko) 2017-04-19 2019-12-20 그릿스톤 온콜로지, 인코포레이티드 신생항원 동정, 제조, 및 용도
CN110720127A (zh) 2017-06-09 2020-01-21 磨石肿瘤生物技术公司 新抗原的鉴别、制造及使用
JP2020532323A (ja) 2017-09-05 2020-11-12 グリットストーン オンコロジー インコーポレイテッド Ｔ細胞療法用の新生抗原の特定法
EP3694532A4 (en) 2017-10-10 2021-07-14 Gritstone Oncology, Inc. IDENTIFICATION OF NEOANTIGENS BY MEANS OF HOT SPOTS
JP2021503897A (ja) 2017-11-22 2021-02-15 グリットストーン オンコロジー インコーポレイテッド 新生抗原のためのジャンクションエピトープ提示の低減
AU2019227813A1 (en) 2018-02-27 2020-10-01 Gritstone Bio, Inc. Neoantigen identification with pan-allele models

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020155447A1 (en) * 2000-04-04 2002-10-24 Maurice Zauderer Gene differentially expressed in breast cancer, and encoded polypeptides
US20030171290A1 (en) * 2000-08-03 2003-09-11 Carr Francis J. Peptides presented by cells
US20160347798A1 (en) * 2014-01-27 2016-12-01 Molecular Templates, Inc. Mhc class i epitope delivering polypeptides and cell-targeted molecules for direct cell killing and immune stimulation via mhc class i presentation and methods regarding the same
US20150278441A1 (en) * 2014-03-25 2015-10-01 Nec Laboratories America, Inc. High-order semi-Restricted Boltzmann Machines and Deep Models for accurate peptide-MHC binding prediction
US20170116371A1 (en) * 2015-10-23 2017-04-27 Google Inc. Neural network for processing aptamer data
US20170221176A1 (en) * 2016-01-29 2017-08-03 Fotonation Limited Convolutional neural network

Cited By (45)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11976326B2 (en) 2013-01-17 2024-05-07 Personalis, Inc. Methods and systems for genetic analysis
US11834718B2 (en) 2013-11-25 2023-12-05 The Broad Institute, Inc. Compositions and methods for diagnosing, evaluating and treating cancer by means of the DNA methylation status
US11725237B2 (en) 2013-12-05 2023-08-15 The Broad Institute Inc. Polymorphic gene typing and somatic change detection using sequencing data
US11452768B2 (en) 2013-12-20 2022-09-27 The Broad Institute, Inc. Combination therapy with neoantigen vaccine
US11939637B2 (en) 2014-12-19 2024-03-26 Massachusetts Institute Of Technology Molecular biomarkers for cancer immunotherapy
US10847252B2 (en) 2015-12-16 2020-11-24 Gritstone Oncology, Inc. Neoantigen identification, manufacture, and use
US11183286B2 (en) 2015-12-16 2021-11-23 Gritstone Bio, Inc. Neoantigen identification, manufacture, and use
US10847253B2 (en) 2015-12-16 2020-11-24 Gritstone Oncology, Inc. Neoantigen identification, manufacture, and use
US11514289B1 (en) * 2016-03-09 2022-11-29 Freenome Holdings, Inc. Generating machine learning models using genetic data
US11952625B2 (en) 2016-05-27 2024-04-09 Personalis, Inc. Methods and systems for genetic analysis
US11643685B2 (en) 2016-05-27 2023-05-09 Personalis, Inc. Methods and systems for genetic analysis
US20190189241A1 (en) * 2016-07-20 2019-06-20 Biontech Rna Pharmaceuticals Gmbh Selecting Neoepitopes as Disease-Specific Targets for Therapy with Enhanced Efficacy
US10769493B2 (en) * 2016-07-27 2020-09-08 Beijing Kuangshi Technology Co., Ltd. Method and apparatus for neural network training and construction and method and apparatus for object detection
US11927591B2 (en) * 2016-08-02 2024-03-12 Georgetown University Methods of identifying novel proteins and antigens in cancer cells
US11965892B2 (en) 2017-02-12 2024-04-23 Biontech Us Inc. HLA-based methods and compositions and uses thereof
US11650211B2 (en) * 2017-02-12 2023-05-16 Biontech Us Inc. HLA-based methods and compositions and uses thereof
US10936949B2 (en) * 2017-02-24 2021-03-02 Deepmind Technologies Limited Training machine learning models using task selection policies to increase learning progress
US11969462B2 (en) * 2017-06-21 2024-04-30 Transgene Personalized vaccine
US20200138923A1 (en) * 2017-06-21 2020-05-07 Transgene Personalized vaccine
CN111433355A (zh) * 2017-08-10 2020-07-17 古德T细胞有限公司 用于癌症治疗的t细胞的活化方法
US11264117B2 (en) 2017-10-10 2022-03-01 Gritstone Bio, Inc. Neoantigen identification using hotspots
US11315016B2 (en) * 2017-10-16 2022-04-26 Illumina, Inc. Deep convolutional neural networks for variant classification
US11798650B2 (en) 2017-10-16 2023-10-24 Illumina, Inc. Semi-supervised learning for training an ensemble of deep convolutional neural networks
US11386324B2 (en) 2017-10-16 2022-07-12 Illumina, Inc. Recurrent neural network-based variant pathogenicity classifier
US11861491B2 (en) 2017-10-16 2024-01-02 Illumina, Inc. Deep learning-based pathogenicity classifier for promoter single nucleotide variants (pSNVs)
CN111630602A (zh) * 2017-11-22 2020-09-04 磨石肿瘤生物技术公司 减少新抗原的接合表位呈递
US11885815B2 (en) 2017-11-22 2024-01-30 Gritstone Bio, Inc. Reducing junction epitope presentation for neoantigens
EP3759131A4 (en) * 2018-02-27 2021-12-01 Gritstone bio, Inc. NEO-ANTIGEN IDENTIFICATION USING PAN-ALLEL MODELS
WO2019168984A1 (en) * 2018-02-27 2019-09-06 Gritstone Oncology, Inc. Neoantigen identification with pan-allele models
WO2019173687A1 (en) * 2018-03-08 2019-09-12 The Trustees Of Indiana University Constrained de novo sequencing of neo-epitope peptides using tandem mass spectrometry
WO2019200398A1 (en) * 2018-04-13 2019-10-17 Dana-Farber Cancer Institute, Inc. Ultra-sensitive detection of cancer by algorithmic analysis
US11814750B2 (en) 2018-05-31 2023-11-14 Personalis, Inc. Compositions, methods and systems for processing or analyzing multi-species nucleic acid samples
US11634767B2 (en) 2018-05-31 2023-04-25 Personalis, Inc. Compositions, methods and systems for processing or analyzing multi-species nucleic acid samples
CN112639136A (zh) * 2018-08-30 2021-04-09 蒙特利尔大学 用于鉴定肿瘤特异性抗原的基于蛋白质基因组学的方法
US11183272B2 (en) * 2018-12-21 2021-11-23 Biontech Us Inc. Method and systems for prediction of HLA class II-specific epitopes and characterization of CD4+ T cells
CN109706065A (zh) * 2018-12-29 2019-05-03 深圳裕策生物科技有限公司 肿瘤新生抗原负荷检测装置及存储介质
CN109712672A (zh) * 2018-12-29 2019-05-03 北京优迅医学检验实验室有限公司 检测基因重排的方法、装置、存储介质及处理器
CN113316818A (zh) * 2019-03-15 2021-08-27 痕准生物科技有限公司 新生抗原的鉴定方法
US11599774B2 (en) * 2019-03-29 2023-03-07 International Business Machines Corporation Training machine learning model
US20200311520A1 (en) * 2019-03-29 2020-10-01 International Business Machines Corporation Training machine learning model
KR102278586B1 (ko) 2020-01-07 2021-07-16 한국과학기술원 신생항원을 스크리닝하는 방법, 시스템 및 그의 용도
KR20210089094A (ko) * 2020-01-07 2021-07-15 한국과학기술원 신생항원을 스크리닝하는 방법, 시스템 및 그의 용도
US20230054552A1 (en) * 2020-04-27 2023-02-23 Flagship Pioneering Innovations Vi, Llc Optimizing Proteins Using Model Based Optimizations
WO2021236854A1 (en) * 2020-05-19 2021-11-25 Gritstone Bio, Inc. Sars-cov-2 vaccines
CN113762416A (zh) * 2021-10-15 2021-12-07 南京澄实生物科技有限公司 基于多模态深度编码的抗原免疫原性预测方法和系统

Also Published As

Publication number Publication date
JP2022133271A (ja) 2022-09-13
KR20180107102A (ko) 2018-10-01
JP2019501967A (ja) 2019-01-24
AU2016369519B2 (en) 2023-04-20
CO2018007417A2 (es) 2018-09-28
BR112018012374A2 (pt) 2018-12-04
IL259931B2 (en) 2024-02-01
US20180330055A1 (en) 2018-11-15
TWI765875B (zh) 2022-06-01
US20210166784A1 (en) 2021-06-03
HK1257865A1 (zh) 2019-11-01
RU2729116C2 (ru) 2020-08-04
IL259931A (en) 2018-07-31
US10847252B2 (en) 2020-11-24
EP4299136A2 (en) 2024-01-03
PH12018501267A1 (en) 2019-02-11
EP3389630A1 (en) 2018-10-24
IL259931B1 (en) 2023-10-01
IL305238A (en) 2023-10-01
US20210098077A1 (en) 2021-04-01
US10847253B2 (en) 2020-11-24
JP7114477B2 (ja) 2022-08-08
EP3389630A4 (en) 2020-01-01
US20190034585A1 (en) 2019-01-31
TW201733598A (zh) 2017-10-01
US20170212984A1 (en) 2017-07-27
SG11201804957VA (en) 2018-07-30
AU2023204618A1 (en) 2023-08-03
EP3389630B1 (en) 2023-11-08
US20190065675A1 (en) 2019-02-28
MX2018007204A (es) 2018-12-11
US11183286B2 (en) 2021-11-23
EP4299136A3 (en) 2024-02-14
AU2016369519A1 (en) 2018-07-12
RU2018124997A3 (ja) 2020-05-29
CA3008641A1 (en) 2017-06-22
PE20181535A1 (es) 2018-09-26
RU2018124997A (ru) 2020-01-16
CN108601731A (zh) 2018-09-28
WO2017106638A1 (en) 2017-06-22
US10055540B2 (en) 2018-08-21

Similar Documents

Publication Publication Date Title
US11183286B2 (en) Neoantigen identification, manufacture, and use
AU2018254526B2 (en) Neoantigen identification, manufacture, and use
AU2018279627B2 (en) Neoantigen identification, manufacture, and use
US11885815B2 (en) Reducing junction epitope presentation for neoantigens
IL273030B1 (en) Neoantigen identification for T-CELL therapy

Legal Events

Date Code Title Description
AS Assignment

Owner name: GRITSTONE ONCOLOGY, INC., CALIFORNIA

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:YELENSKY, ROMAN;DERTI, ADNAN;BULIK-SULLIVAN, BRENDAN;AND OTHERS;SIGNING DATES FROM 20170515 TO 20170517;REEL/FRAME:042445/0268

STPP Information on status: patent application and granting procedure in general

Free format text: RESPONSE TO NON-FINAL OFFICE ACTION ENTERED AND FORWARDED TO EXAMINER

STPP Information on status: patent application and granting procedure in general

Free format text: NON FINAL ACTION MAILED

STPP Information on status: patent application and granting procedure in general

Free format text: RESPONSE TO NON-FINAL OFFICE ACTION ENTERED AND FORWARDED TO EXAMINER

STPP Information on status: patent application and granting procedure in general

Free format text: FINAL REJECTION MAILED

STPP Information on status: patent application and granting procedure in general

Free format text: DOCKETED NEW CASE - READY FOR EXAMINATION

STCB Information on status: application discontinuation

Free format text: ABANDONED -- FAILURE TO RESPOND TO AN OFFICE ACTION

AS Assignment

Owner name: GRITSTONE BIO, INC., CALIFORNIA

Free format text: CHANGE OF NAME;ASSIGNOR:GRITSTONE ONCOLOGY, INC.;REEL/FRAME:056171/0666

Effective date: 20210503