WO2021003470A1 - Approches de décodage pour l'identification de protéines et de peptides - Google Patents

Approches de décodage pour l'identification de protéines et de peptides Download PDF

Info

Publication number
WO2021003470A1
WO2021003470A1 PCT/US2020/040829 US2020040829W WO2021003470A1 WO 2021003470 A1 WO2021003470 A1 WO 2021003470A1 US 2020040829 W US2020040829 W US 2020040829W WO 2021003470 A1 WO2021003470 A1 WO 2021003470A1
Authority
WO
WIPO (PCT)
Prior art keywords
peptide
protein
peptides
sample
candidate
Prior art date
Application number
PCT/US2020/040829
Other languages
English (en)
Inventor
Sujal M. Patel
Parag Mallick
Jarrett D. EGERTSON
Original Assignee
Nautilus Biotechnology, 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 Nautilus Biotechnology, Inc. filed Critical Nautilus Biotechnology, Inc.
Publication of WO2021003470A1 publication Critical patent/WO2021003470A1/fr

Links

Classifications

    • 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
    • G16B35/00—ICT specially adapted for in silico combinatorial libraries of nucleic acids, proteins or peptides
    • G16B35/20—Screening of libraries
    • C—CHEMISTRY; METALLURGY
    • C40—COMBINATORIAL TECHNOLOGY
    • C40B—COMBINATORIAL CHEMISTRY; LIBRARIES, e.g. CHEMICAL LIBRARIES
    • C40B30/00—Methods of screening libraries
    • C40B30/04—Methods of screening libraries by measuring the ability to specifically bind a target molecule, e.g. antibody-antigen binding, receptor-ligand binding
    • 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
    • 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

Definitions

  • a sample of unknown proteins or peptides may be exposed to individual affinity reagent probes, pooled affinity reagent probes, or a combination of individual affinity reagent probes and pooled affinity reagent probes.
  • the identification may comprise estimation of a confidence level that each of one or more candidate proteins or peptides is present in the sample.
  • peptides or proteins may be structurally modified to enable detection.
  • structural modification may be targeted to a single protein or peptide, many proteins or peptides, or a specific chemical modality present in all or a subset of the proteins or peptides being measured (e.g.
  • the empirical measurements may comprise binding of reagents that are specific for a terminus of a protein or peptide (e.g., an N-terminus or a C-terminus).
  • the empirical measurements may comprise detected signals indicative of amino acid information (e.g., detected by optical detection, spectroscopic detection, electrostatic detection, electrochemical detection, magnetic detection, fluorescence detection, surface plasmon resonance (SPR), and the like) of a protein or peptide.
  • amino acids of a protein or peptide may be selectively modified (e.g., chemically modified or fluorescently labeled) such that a signal may be measured by a detector suitable to detect the modification.
  • amino acids of a protein or peptide may be directly identified or measured by a detector (e.g., a protein nanopore).
  • the empirical measurements may comprise differential measurements obtained following a perturbation of the proteins or peptides (e.g., a change in conditions of the proteins or peptides such as pH, and/or performing an enzymatic reaction of the proteins or peptides).
  • Methods and systems provided herein may comprise algorithms for identifying proteins and peptides based on a sequence of experiments performed on fully-intact proteins, protein fragments, or peptides.
  • Each experiment may be an empirical measurement performed on a protein or peptide, and may provide information which may be useful for identifying the protein or peptide.
  • Examples of experiments include measurement of the binding of an affinity reagent (e.g., antibody or aptamer), length of a protein or peptide, hydrophobicity of a protein or peptide, and isoelectric point of a protein or peptide.
  • Further examples of experiments include selective modification of proteins or peptides by a single chemical reaction or a sequence of chemical reactions followed by empirical measurement performed on the proteins or peptides.
  • the empirical measurements may comprise information on the chemical modification (e.g., if there was a modification, degree of modification, type of modification) of each protein or peptide resulting from the chemical reaction or series of chemical reactions.
  • Individual experimental outcomes need not comprise information sufficient to identify proteins.
  • multiple experiments, each providing partial information for identification may be performed on the same protein or peptide substrates in succession to generate combined information sufficient to generate a protein or peptide identification.
  • the empirical measurements may comprise incomplete amino acid information of proteins or peptides (e.g., amino acid information of only a subset of the protein or peptide).
  • Such incomplete amino acid information may include amino acid information of only a subset of the amino acids (e.g., a set of chemically modified amino acids) in a protein or peptide.
  • such incomplete amino acid information may include a presence or absence of amino acids in the protein or peptide, a count of amino acids in the protein or peptide, an order of amino acids in the protein or peptide, or a combination thereof.
  • Information about experimental outcomes may be used to calculate probabilities or likelihoods of protein or peptide candidates, and/or to infer identity of a protein or peptide by selecting the protein or peptide from a list of protein or peptide candidates that maximizes the likelihood of the observed experimental outcomes.
  • Methods and systems provided herein may also comprise a collection of protein or peptide candidates, and algorithms to calculate the probability of experimental outcomes from each of these protein or peptide candidates.
  • the present disclosure provides a computer-implemented method for identifying a peptide in a sample of unknown peptides, the method comprising: (a) receiving, by said computer, information of a plurality of empirical measurements performed on said unknown peptides in said sample; (b) comparing, by said computer, at least a portion of said information of said plurality of said empirical measurements against a database comprising a plurality of peptide sequences, each peptide sequence corresponding to a candidate peptide among a plurality of candidate peptides; and (c) for each of one or more candidate peptides in said plurality of candidate peptides, generating, by said computer, one or more of: (i) a probability that said candidate peptide generates said information of said plurality of empirical measurements, (ii) a probability that said plurality of empirical measurements is not observed given that said candidate peptide is present in said sample, and (iii) a probability that said candidate peptide is present in said sample
  • two or more of said plurality of empirical measurements are selected from the group consisting of: (i) binding measurements of each of one or more affinity reagent probes to said unknown peptides in said sample, each affinity reagent probe configured to selectively bind to one or more candidate peptides among said plurality of candidate peptides; (ii) length of one or more of said unknown peptides in said sample; (iii) hydrophobicity of one or more of said unknown peptides in said sample; (iv) isoelectric point of one or more of said unknown peptides in said sample; (v) a presence or absence of each of a plurality of amino acids in one or more of said unknown peptides in said sample; (vi) a count of each of a plurality of amino acids in one or more of said unknown peptides in said sample; (vii) an order of each of a plurality of amino acids in one or more of said unknown peptides in said sample; (viii) N-
  • generating said plurality of probabilities further comprises receiving additional information of binding measurements of each of a plurality of additional affinity reagent probes, each additional affinity reagent probe configured to selectively bind to one or more candidate peptides among said plurality of candidate peptides.
  • the method further comprises generating, for said each of one or more candidate peptides, a confidence level that said candidate peptide matches one of said unknown peptides in said sample.
  • said plurality of affinity reagent probes comprises no more than about 50 affinity reagent probes. In some embodiments, said plurality of affinity reagent probes comprises no more than about 100 affinity reagent probes. In some embodiments, said plurality of affinity reagent probes comprises no more than about 200 affinity reagent probes. In some embodiments, said plurality of affinity reagent probes comprises no more than about 300 affinity reagent probes. In some embodiments, said plurality of affinity reagent probes comprises no more than about 500 affinity reagent probes. In some embodiments, said plurality of affinity reagent probes comprises more than about 500 affinity reagent probes. In some embodiments, the method further comprises generating a paper or electronic report identifying said peptides in said sample.
  • said sample comprises a biological sample.
  • said biological sample is obtained from a subject.
  • the method further comprises identifying a disease state in said subject based at least on said plurality of probabilities.
  • (c) comprises, for each of one or more candidate peptides in said plurality of candidate peptides, generating, by said computer, (i) said probability that said candidate peptide generates said information of said plurality of empirical measurements. In some embodiments, (c) comprises, for each of one or more candidate peptides in said plurality of candidate peptides, generating, by said computer, (ii) said probability that said plurality of empirical measurements is not observed given that said candidate peptide is present in said sample. In some embodiments, (c) comprises, for each of one or more candidate peptides in said plurality of candidate peptides, generating, by said computer, (iii) said probability that said candidate peptide is present in said sample.
  • said measurement outcome comprises binding of affinity reagent probes. In some embodiments, said measurement outcome comprises non-specific binding of affinity reagent probes. In some embodiments, said measurement outcome comprises binding of affinity reagent probes. In some embodiments, said measurement outcome comprises non-specific binding of affinity reagent probes. In some embodiments, said empirical measurements comprise binding of affinity reagent probes. In some embodiments, said empirical measurements comprise non-specific binding of affinity reagent probes.
  • the method further comprises generating a sensitivity of peptide identification with a pre-determined threshold.
  • said pre-determined threshold is less than about 1%, less than about 2%, less than about 3%, less than about 4%, less than about 5%, less than about 6%, less than about 7%, less than about 8%, less than about 9%, or less than about 10% of being incorrect.
  • the pre-determined threshold may correspond to a given false detection rate (FDR) of identifying substrates, which may be less than about 1%, less than about 2%, less than about 3%, less than about 4%, less than about 5%, less than about 6%, less than about 7%, less than about 8%, less than about 9%, or less than about 10%.
  • FDR false detection rate
  • said peptide in said sample is truncated or degraded.
  • said peptide in said sample does not originate from a protein terminus.
  • said empirical measurements comprise length of one or more of said unknown peptides in said sample. In some embodiments, said empirical measurements comprise hydrophobicity of one or more of said unknown peptides in said sample. In some embodiments, said empirical measurements comprise isoelectric point of one or more of said unknown peptides in said sample. In some embodiments, said empirical measurements comprise a presence or absence of each of a plurality of amino acids in one or more of said unknown peptides in said sample, a count of each of a plurality of amino acids in one or more of said unknown peptides in said sample, an order of each of a plurality of amino acids in one or more of said unknown peptides in said sample, or a combination thereof.
  • said empirical measurements comprise measurements performed on mixtures of antibodies. In some embodiments, said empirical measurements comprise measurements performed on samples obtained from a plurality of species. In some embodiments, said empirical measurements comprise measurements performed on samples in the presence of single amino acid variants (SAVs) caused by non-synonymous single-nucleotide polymorphisms (SNPs). [0015] In some embodiments, one or more of said plurality of empirical measurements comprise incomplete amino acid information of a plurality of amino acids in one or more of said unknown peptides in said sample.
  • SAVs single amino acid variants
  • SNPs non-synonymous single-nucleotide polymorphisms
  • said one or more of said plurality of empirical measurements comprise a presence or absence of each of said plurality of amino acids, a count of each of said plurality of amino acids, an order of each of said plurality of amino acids, or a combination thereof. In some embodiments, said one or more of said plurality of empirical measurements comprise two or more of: a presence or absence of each of said plurality of amino acids, a count of each of said plurality of amino acids, and an order of each of said plurality of amino acids. In some embodiments, said one or more of said plurality of empirical measurements.
  • measurements comprise a presence or absence of each of said plurality of amino acids, a count of each of said plurality of amino acids, and an order of each of said plurality of amino acids.
  • the present disclosure provides a system, comprising a controller comprising or capable of accessing, a non-transitory computer-readable medium comprising machine-executable instructions which, upon execution by one or more computer processors, perform a method for identifying a peptide in a sample of unknown peptides, said method comprising: (a) receiving information of a plurality of empirical measurements performed on said unknown peptides in said sample; (b) comparing at least a portion of said information of said plurality of said empirical measurements against a database comprising a plurality of peptide sequences, each peptide sequence corresponding to a candidate peptide among a plurality of candidate peptides; and (c) for each of one or more candidate peptides in said plurality of candidate peptides, generating one or more of: (i) a probability that said candidate peptide generates said information of said plurality of empirical measurements, (ii) a probability that said plurality of empirical measurements is not observed given that said candidate
  • two or more of said plurality of empirical measurements are selected from the group consisting of: (i) binding measurements of each of one or more affinity reagent probes to said unknown peptides in said sample, each affinity reagent probe configured to selectively bind to one or more candidate peptides among said plurality of candidate peptides; (ii) length of one or more of said unknown peptides in said sample; (iii) hydrophobicity of one or more of said unknown peptides in said sample; (iv) isoelectric point of one or more of said unknown peptides in said sample; (v) a presence or absence of each of a plurality of amino acids in one or more of said unknown peptides in said sample; (vi) a count of each of a plurality of amino acids in one or more of said unknown peptides in said sample; (vii) an order of each of a plurality of amino acids in one or more of said unknown peptides in said sample; (viii) N-
  • generating said plurality of probabilities further comprises receiving additional information of binding measurements of each of a plurality of additional affinity reagent probes, each additional affinity reagent probe configured to selectively bind to one or more candidate peptides among said plurality of candidate peptides.
  • said method further comprises generating, for said each of one or more candidate peptides, a confidence level that said candidate peptide matches one of said unknown peptides in said sample.
  • said plurality of affinity reagent probes comprises no more than about 50 affinity reagent probes. In some embodiments, said plurality of affinity reagent probes comprises no more than about 100 affinity reagent probes. In some embodiments, said plurality of affinity reagent probes comprises no more than about 200 affinity reagent probes. In some embodiments, said plurality of affinity reagent probes comprises no more than about 300 affinity reagent probes. In some embodiments, said plurality of affinity reagent probes comprises no more than about 500 affinity reagent probes. In some embodiments, said plurality of affinity reagent probes comprises more than about 500 affinity reagent probes. In some embodiments, said method further comprises generating a paper or electronic report identifying said peptides in said sample.
  • said sample comprises a biological sample.
  • said biological sample is obtained from a subject.
  • the method further comprises identifying a disease state in said subject based at least on said plurality of probabilities.
  • (c) comprises, for each of one or more candidate peptides in said plurality of candidate peptides, generating, by said computer, (i) said probability that said candidate peptide generates said information of said plurality of empirical measurements. In some embodiments, (c) comprises, for each of one or more candidate peptides in said plurality of candidate peptides, generating, by said computer, (ii) said probability that said plurality of empirical measurements is not observed given that said candidate peptide is present in said sample. In some embodiments, (c) comprises, for each of one or more candidate peptides in said plurality of candidate peptides, generating, by said computer, (iii) said probability that said candidate peptide is present in said sample.
  • said measurement outcome comprises binding of affinity reagent probes. In some embodiments, said measurement outcome comprises non-specific binding of affinity reagent probes. In some embodiments, said measurement outcome comprises binding of affinity reagent probes. In some embodiments, said measurement outcome comprises non-specific binding of affinity reagent probes. In some embodiments, said empirical measurements comprise binding of affinity reagent probes. In some embodiments, said empirical measurements comprise non-specific binding of affinity reagent probes.
  • said method further comprises generating a sensitivity of peptide identification with a pre-determined threshold.
  • said pre-determined threshold is less than about 1%, less than about 2%, less than about 3%, less than about 4%, less than about 5%, less than about 6%, less than about 7%, less than about 8%, less than about 9%, or less than about 10% of being incorrect.
  • the pre-determined threshold may correspond to a given false detection rate (FDR) of identifying substrates, which may be less than about 1%, less than about 2%, less than about 3%, less than about 4%, less than about 5%, less than about 6%, less than about 7%, less than about 8%, less than about 9%, or less than about 10%.
  • FDR false detection rate
  • said peptide in said sample is truncated or degraded.
  • said peptide in said sample does not originate from a protein terminus.
  • said empirical measurements comprise length of one or more of said unknown peptides in said sample. In some embodiments, said empirical measurements comprise hydrophobicity of one or more of said unknown peptides in said sample. In some embodiments, said empirical measurements comprise isoelectric point of one or more of said unknown peptides in said sample. In some embodiments, said empirical measurements comprise a presence or absence of each of a plurality of amino acids in one or more of said unknown peptides in said sample, a count of each of a plurality of amino acids in one or more of said unknown peptides in said sample, an order of each of a plurality of amino acids in one or more of said unknown peptides in said sample, or a combination thereof.
  • said empirical measurements comprise measurements performed on mixtures of antibodies. In some embodiments, said empirical measurements comprise measurements performed on samples obtained from a plurality of species. In some embodiments, said empirical measurements comprise measurements performed on samples in the presence of single amino acid variants (SAVs) caused by non-synonymous single-nucleotide polymorphisms (SNPs).
  • SAVs single amino acid variants
  • SNPs non-synonymous single-nucleotide polymorphisms
  • one or more of said plurality of empirical measurements comprise incomplete amino acid information of a plurality of amino acids in one or more of said unknown peptides in said sample.
  • said one or more of said plurality of empirical measurements comprise a presence or absence of each of said plurality of amino acids, a count of each of said plurality of amino acids, an order of each of said plurality of amino acids, or a combination thereof.
  • said one or more of said plurality of empirical measurements comprise two or more of: a presence or absence of each of said plurality of amino acids, a count of each of said plurality of amino acids, and an order of each of said plurality of amino acids.
  • said one or more of said plurality of empirical measurements comprise incomplete amino acid information of a plurality of amino acids in one or more of said unknown peptides in said sample.
  • said one or more of said plurality of empirical measurements comprise a presence or absence of each of said plurality of amino acids, a count of each of said plurality of amino acids, an order of each of said plurality of amino acids, or
  • measurements comprise a presence or absence of each of said plurality of amino acids, a count of each of said plurality of amino acids, and an order of each of said plurality of amino acids.
  • Another aspect of the present disclosure provides a non-transitory computer readable medium comprising machine executable code that, upon execution by one or more computer processors, implements any of the methods above or elsewhere herein.
  • Another aspect of the present disclosure provides a system comprising one or more computer processors and computer memory coupled thereto.
  • the computer memory comprises machine executable code that, upon execution by the one or more computer processors, implements any of the methods above or elsewhere herein.
  • FIG. 1 illustrates an example flowchart of protein or peptide identification of unknown proteins in a biological sample, in accordance with disclosed embodiments.
  • FIG. 2 illustrates the sensitivity of affinity reagent probes (e.g., the percent of substrates identified with a false detection rate (FDR) of less than 1%) plotted against the number of probe recognition sites (e.g., trimer-binding epitopes) in the affinity reagent probe (ranging up to 100 probe recognition sites or trimer-binding epitopes), for three different experimental cases (with 50, 100, and 200 probes used, as denoted by the gray, black, and white circles, respectively), in accordance with disclosed embodiments.
  • FDR false detection rate
  • FIG. 3 illustrates the sensitivity of affinity reagent probes (e.g., the percent of substrates identified with a false detection rate (FDR) of less than 1%) plotted against the number of probe recognition sites (e.g., trimer-binding epitopes)in the affinity reagent probe (ranging up to 700 probe recognition sites or trimer-binding epitopes) for three different experimental cases (with 50, 100, and 200 probes used, as denoted by the gray, black, and white circles, respectively), in accordance with disclosed embodiments.
  • FDR false detection rate
  • FIG. 4 illustrates plots showing the sensitivity of protein or peptide identification with experiments using 100 (left), 200 (center), or 300 probes (right), in accordance with disclosed embodiments.
  • FIG. 5 illustrates plots showing the sensitivity of protein or peptide identification with experiments using various protein fragmentation approaches.
  • protein identification performance is shown with 50, 100, 200, and 300 affinity reagent measurements (in the 4 panels from left to right), with maximum fragment length values of 50, 100, 200, 300, 400, and 500 (as denoted by the hexagons, down-pointing triangles, up- pointing triangles, diamonds, rectangles, and circles, respectively), in accordance with disclosed embodiments.
  • FIG. 6 illustrates plots showing the sensitivity of identification of human proteins (percent of substrates identified at an FDR of less than 1%) with experiments using various combinations of types of measurements), in accordance with disclosed embodiments.
  • FIG. 7 illustrates plots showing the sensitivity of protein or peptide identification with experiments using 50, 100, 200, or 300 affinity reagent probe passes against unknown proteins from either E. coli , yeast, or human (as denoted by the circles, triangles, and squares, respectively), in accordance with disclosed embodiments.
  • FIG. 8 illustrates a plot showing the binding probability (y-axis, left) and sensitivity of protein or peptide identification (y-axis, right) against iteration (x-axis), in accordance with disclosed embodiments.
  • FIG. 9 shows a comparison of the estimated false identification rate to the true false identification rate for a simulated 200-probe experiment demonstrates accurate false
  • FIG. 10 illustrates a computer system that is programmed or otherwise configured to implement methods provided herein.
  • FIG. 11 illustrates the performance of a censored protein or peptide identification vs. an uncensored protein or peptide identification approach.
  • FIG. 12 illustrates the tolerance of censored protein or peptide identification and uncensored protein or peptide identification approaches to random“false negative” binding outcomes.
  • FIG. 13 illustrates the tolerance of censored protein or peptide identification and uncensored protein or peptide identification approaches to random“false positive” binding outcomes.
  • FIG. 14 illustrates the performance of censored protein or peptide identification and uncensored protein or peptide identification approaches with overestimated or underestimated affinity reagent binding probabilities.
  • FIG. 15 illustrates the performance of censored protein or peptide identification and uncensored protein or peptide identification approaches using affinity reagents with unknown binding epitopes.
  • FIG. 16 illustrates the performance of censored protein or peptide identification and uncensored protein or peptide identification approaches using affinity reagents with missing binding epitopes.
  • FIG. 17 illustrates the performance of censored protein or peptide identification and uncensored protein or peptide identification approaches using affinity reagents targeting the top 300 most abundant trimers in the proteome, 300 randomly selected trimers in the proteome, or the 300 least abundant trimers in the proteome.
  • FIG. 18 illustrates the performance of censored protein or peptide identification and uncensored protein or peptide identification approaches using affinity reagents with random or biosimilar off-target sites.
  • FIG. 19 illustrates the performance of censored protein or peptide identification and uncensored protein or peptide identification approaches using a set of optimal affinity reagents (probes).
  • FIG. 20 illustrates the performance of censored protein or peptide identification and uncensored protein or peptide identification approaches using unmixed candidate affinity reagents and mixtures of candidate affinity reagents.
  • FIG. 21 illustrates two hybridization steps in reinforcing a binding between an affinity reagent and a protein, in accordance with some embodiments.
  • FIG. 22 illustrates the performance of protein or peptide identification using a collection of reagents for selective modification and detection of 4 amino acids (K, D, C, and W), in accordance with some embodiments.
  • FIG. 23 illustrates the performance of protein or peptide identification using a collection of reagents for selective modification and detection of 20 amino acids (R, H, K, D, E, S, T, N, Q, C, G, P, A, V, I, L, M, F, Y, and W), in accordance with some embodiments.
  • FIG. 24 illustrates the performance of protein or peptide identification using measurements of order of amino acids, where all amino acids are measured with a detection probability (equal to reaction efficiency) indicated on the x-axis, and the y-axis indicates the percent of proteins in the sample identified with a false discovery rate below 1%, in accordance with some embodiments.
  • sample generally refers to a biological sample (e.g., a sample containing protein).
  • the samples may be taken from tissue or cells or from the environment of tissue or cells.
  • the sample may comprise, or be derived from, a tissue biopsy, blood, blood plasma, extracellular fluid, dried blood spots, cultured cells, culture media, discarded tissue, plant matter, synthetic proteins, bacterial and/or viral samples, fungal tissue, archaea, or protozoans.
  • the sample may have been isolated from the source prior to collection.
  • Samples may comprise forensic evidence. Non-limiting examples include a fingerprint, saliva, urine, blood, stool, semen, or other bodily fluids isolated from the primary source prior to collection.
  • one or more proteins or peptides are isolated from their primary source (cells, tissue, bodily fluids such as blood, environmental samples, etc.) during sample preparation.
  • the sample may be derived from an extinct species including, but not limited to, samples derived from fossils.
  • the proteins or peptides may or may not be purified or otherwise enriched from their primary source. In some cases, the primary source is homogenized prior to further processing. In some cases, cells are lysed using a buffer such as RIPA buffer. Denaturing buffers may also be used at this stage.
  • the sample may be filtered or centrifuged to remove lipids and particulate matter.
  • the sample may also be purified to remove nucleic acids, or may be treated with RNases and DNases.
  • the sample may contain intact proteins, denatured proteins, protein fragments, partially degraded proteins, intact peptides, denatured peptides, peptide fragments, or partially degraded peptides.
  • the sample may be taken from a subject with a disease or disorder.
  • the disease or disorder may be an infectious disease, an immune disorder or disease, a cancer, a genetic disease, a degenerative disease, a lifestyle disease, an injury, a rare disease, or an age related disease.
  • the infectious disease may be caused by bacteria, viruses, fungi, and/or parasites.
  • Non-limiting examples of cancers include Bladder cancer, Lung cancer, Brain cancer, Melanoma, Breast cancer, Non-Hodgkin lymphoma, Cervical cancer, Ovarian cancer, Colorectal cancer, Pancreatic cancer, Esophageal cancer, Prostate cancer, Kidney cancer, Skin cancer, Leukemia, Thyroid cancer, Liver cancer, and Uterine cancer.
  • genetic diseases or disorders include, but are not limited to, multiple sclerosis (MS), cystic fibrosis, Charcot-Marie-Tooth disease, Huntington's disease, Koz-Jeghers syndrome, Down syndrome, Rheumatoid arthritis, and Tay-Sachs disease.
  • Non-limiting examples of lifestyle diseases include obesity, diabetes, arteriosclerosis, heart disease, stroke, hypertension, liver cirrhosis, nephritis, cancer, chronic obstructive pulmonary disease (COPD), hearing problems, and chronic backache.
  • MS multiple sclerosis
  • cystic fibrosis Charcot-Marie-Tooth disease
  • Huntington's disease Huntington's disease
  • Peutz-Jeghers syndrome Down syndrome
  • Rheumatoid arthritis and Tay-Sachs disease.
  • Non-limiting examples of lifestyle diseases include obesity, diabetes, arteriosclerosis, heart disease, stroke, hypertension, liver cirrhosis, nephritis, cancer, chronic obstructive pulmonary
  • injuries include, but are not limited to, abrasion, brain injuries, bruising, burns, concussions, congestive heart failure, construction injuries, dislocation, flail chest, fracture, hemothorax, herniated disc, hip pointer, hypothermia, lacerations, pinched nerve, pneumothorax, rib fracture, sciatica, spinal cord injury, tendons ligaments fascia injury, traumatic brain injury, and whiplash.
  • the sample may be taken before and/or after treatment of a subject with a disease or disorder. Samples may be taken before and/or after a treatment. Samples may be taken during a treatment or a treatment regime. Multiple samples may be taken from a subject to monitor the effects of the treatment over time. The sample may be taken from a subject known or suspected of having an infectious disease for which diagnostic antibodies are not available.
  • the sample may be taken from a subject suspected of having a disease or a disorder.
  • the sample may be taken from a subject experiencing unexplained symptoms, such as fatigue, nausea, weight loss, aches and pains, weakness, or memory loss.
  • the sample may be taken from a subject having explained symptoms.
  • the sample may be taken from a subject at risk of developing a disease or disorder due to factors such as familial history, age, environmental exposure, lifestyle risk factors, or presence of other known risk factors.
  • the sample may be taken from an embryo, fetus, or pregnant woman.
  • the sample may comprise of proteins isolated from the mother’s blood plasma.
  • proteins isolated from circulating fetal cells in the mother’s blood are proteins isolated from circulating fetal cells in the mother’s blood.
  • the sample may be taken from a healthy individual.
  • samples may be taken longitudinally from the same individual.
  • samples acquired longitudinally may be analyzed with the goal of monitoring individual health and early detection of health issues.
  • the sample may be collected at a home setting or at a point-of-care setting and subsequently transported by a mail delivery, courier delivery, or other transport method prior to analysis.
  • a home user may collect a blood spot sample through a finger prick, which blood spot sample may be dried and subsequently transported by mail delivery prior to analysis.
  • samples acquired longitudinally may be used to monitor response to stimuli expected to impact healthy, athletic performance, or cognitive performance. Non-limiting examples include response to medication, dieting, or an exercise regimen.
  • Proteins or peptides of the sample may be treated to remove modifications that may interfere with epitope binding.
  • the proteins or peptides may be enzymatically treated.
  • the proteins or peptides may be glycosidase treated to remove post- translational glycosylation.
  • the proteins or peptides may be treated with a reducing agent to reduce disulfide binds within the protein.
  • the proteins or peptides may be treated with a phosphatase to remove phosphate groups.
  • post-translational modifications include acetate, amide groups, methyl groups, lipids, ubiquitin, myristoylation, palmitoylation, isoprenylation or prenylation (e.g., farnesol and geranylgeraniol), farnesylation, geranylgeranylation, glypiation, lipoylation, flavin moiety attachment, phosphopantetheinylation, and retinylidene Schiff base formation.
  • prenylation e.g., farnesol and geranylgeraniol
  • Proteins or peptides of the sample may be treated by modifying one or more residues to make them more amenable to being bound or detected.
  • the proteins or peptides may be bound by or detected by an affinity reagent.
  • the proteins or peptides may be bound by reagents that are specific for a terminus of a protein or peptide (e.g., an N-terminus or a C-terminus).
  • signals indicative of amino acid information e.g., detected by optical detection, spectroscopic detection, electrostatic detection, electrochemical detection, magnetic detection, fluorescence detection, surface plasmon resonance (SPR), and the like
  • SPR surface plasmon resonance
  • amino acids of a protein or peptide may be selectively modified (e.g., chemically modified or fluorescently labeled) such that a signal may be measured by a detector suitable to detect the modification.
  • amino acids of a protein or peptide may be directly identified or measured by a detector (e.g., a protein nanopore).
  • the empirical measurements may comprise differential measurements obtained following a perturbation of the proteins or peptides (e.g., a change in conditions of the proteins or peptides such as pH, and/or performing an enzymatic reaction of the proteins or peptides).
  • proteins or peptides of the sample may be treated to retain post-translational protein modifications that may facilitate or enhance epitope binding.
  • phosphatase inhibitors may be added to the sample.
  • oxidizing agents may be added to protect disulfide bonds.
  • Proteins or peptides of the sample may be denatured in full or in part.
  • proteins or peptides can be fully denatured.
  • Proteins or peptides may be denatured by application of an external stress such as a detergent, a strong acid or base, a concentrated inorganic salt, an organic solvent (e.g., alcohol or chloroform), radiation, or heat.
  • an external stress such as a detergent, a strong acid or base, a concentrated inorganic salt, an organic solvent (e.g., alcohol or chloroform), radiation, or heat.
  • Proteins or peptides may be denatured by addition of a denaturing buffer. Proteins or peptides may also be precipitated, lyophilized, and suspended in denaturing buffer. Proteins or peptides may be denatured by heating. Methods of denaturing that are unlikely to cause chemical modifications to the proteins or peptides may be preferred.
  • Proteins or peptides of the sample may be treated to produce shorter polypeptides, either before or after conjugation. Remaining proteins or peptides may be partially digested with an enzyme such as ProteinaseK to generate protein or peptide fragments or may be left intact. In further examples, the proteins or peptides may be exposed to proteases such as trypsin. For example, a protein may be subjected to trypsin digestion to produce a plurality of tryptic peptides, each of which can be subjected to peptide inference as described herein.
  • proteases may include serine proteases, cysteine proteases, threonine proteases, aspartic proteases, glutamic proteases, metalloproteases, and asparagine peptide lyases.
  • treatment of proteins or peptides may comprise subjecting the proteins or peptides to Edman processes (e.g., Edman degradation), whereby one residue at a time is sequentially labeled and removed by cleaving from the amino end of a protein or peptide without disrupting peptide bonds between other amino acid residues.
  • extremely large proteins or peptides may include proteins or peptides that are at least about 400 kilodalton (kD), 450 kD, 500 kD, 600 kD, 650 kD, 700 kD, 750 kD, 800 kD, or 850 kD.
  • extremely large proteins or peptides may include proteins or peptides that are at least about 8,000 amino acids, about 8,500 amino acids, about 9,000 amino acids, about 9,500 amino acids, about 10,000 amino acids, about 10,500 amino acids, about 11,000 amino acids, or about 15,000 amino acids.
  • small proteins or peptides may include proteins or peptides that are less than about 10 kD, 9 kD,
  • small proteins or peptides may include proteins or peptides that are less than about 50 amino acids, 45 amino acids, 40 amino acids, 35 amino acids, or about 30 amino acids. Extremely large or small proteins or peptides can be removed by size exclusion chromatography. Extremely large proteins or peptides may be isolated by size exclusion chromatography, treated with proteases to produce moderately sized polypeptides, and recombined with the moderately size proteins or peptides of the sample.
  • Proteins or peptides of the sample may be tagged, e.g., with identifiable tags, to allow for multiplexing of samples.
  • identifiable tags include:
  • Fluorophores may include fluorescent proteins such as GFP, YFP, RFP, eGFP, mCherry, tdtomato, FITC, Alexa Fluor 350, Alexa Fluor 405, Alexa Fluor 488, Alexa Fluor 532, Alexa Fluor 546, Alexa Fluor 555, Alexa Fluor 568, Alexa Fluor 594, Alexa Fluor 647, Alexa Fluor 680, Alexa Fluor 750, Pacific Blue, Coumarin, BODIPY FL, Pacific Green, Oregon Green, Cy3, Cy5, Pacific Orange, TRITC, Texas Red, Phycoerythrin, and Allophcocyanin.
  • fluorescent proteins such as GFP, YFP, RFP, eGFP, mCherry, tdtomato, FITC, Alexa Fluor 350, Alexa Fluor 405, Alexa Fluor 488, Alexa Fluor 532, Alexa Fluor 546, Alexa Fluor 555, Alexa Fluor 568, Alexa Fluor 594,
  • any number of protein or peptides samples may be multiplexed.
  • a multiplexed reaction may contain proteins or peptides from 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, about 20, about 25, about 30, about 35, about 40, about 45, about 50, about 55, about 60, about 65, about 70, about 75, about 80, about 85, about 90, about 95, about 100, or more than about 100 initial samples.
  • the identifiable tags may provide a way to interrogate each protein or peptide as to its sample of origin, or may direct proteins or peptides from different samples to segregate to different areas or a solid support.
  • the proteins or peptides are then applied to a functionalized substrate to chemically attach proteins or peptides to the substrate.
  • any number of protein or peptide samples may be mixed prior to analysis without tagging or multiplexing.
  • a multiplexed reaction may contain proteins or peptides from 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, about 20, about 25, about 30, about 35, about 40, about 45, about 50, about 55, about 60, about 65, about 70, about 75, about 80, about 85, about 90, about 95, about 100, or more than about 100 initial samples.
  • diagnostics for rare conditions may be performed on pooled samples. Analysis of individual samples may then be performed only from samples in a pool that tested positive for the diagnostic. Samples may be multiplexed without tagging using a combinatorial pooling design in which samples are mixed into pools in a manner that allows signal from individual samples to be resolved from the analyzed pools using computational demultiplexing.
  • substrate generally refers to a substrate capable of forming a solid support.
  • Substrates, or solid substrates can refer to any solid surface to which proteins can be covalently or non-covalently attached.
  • Non-limiting examples of solid substrates include particles, beads, slides, surfaces of elements of devices, membranes, flow cells, wells, chambers, macrofluidic chambers, microfluidic chambers, channels, microfluidic channels, or any other surfaces.
  • Substrate surfaces can be flat or curved, or can have other shapes, and can be smooth or textured. Substrate surfaces may contain microwells.
  • the substrate can be composed of glass, carbohydrates such as dextrans, plastics such as polystyrene or polypropylene, polyacrylamide, latex, silicon, metals such as gold, or cellulose, and may be further modified to allow or enhance covalent or non-covalent attachment of the proteins or peptides.
  • the substrate surface may be functionalized by modification with specific functional groups, such as maleic or succinic moieties, or derivatized by modification with a chemically reactive group, such as amino, thiol, or acrylate groups, such as by silanization.
  • Suitable silane reagents include aminopropyltrimethoxysilane, aminopropyltriethoxysilane and 4-aminobutyltriethoxysilane.
  • the substrate may be functionalized with N-Hydroxysuccinimide (NHS) functional groups. Glass surfaces can also be derivatized with other reactive groups, such as acrylate or epoxy, using, e.g., epoxysilane, acrylatesilane or acrylamidesilane.
  • the substrate and process for protein or peptide attachment are preferably stable for repeated binding, washing, imaging and eluting steps.
  • the substrate may be a slide, a flow cell, or a microscaled or nanoscaled structure (e.g., an ordered structure such as microwells, micropillars, single molecule arrays, nanoballs, nanopillars, or nanowires).
  • a microscaled or nanoscaled structure e.g., an ordered structure such as microwells, micropillars, single molecule arrays, nanoballs, nanopillars, or nanowires.
  • the spacing of the functional groups on the substrate may be ordered or random.
  • An ordered array of functional groups may be created by, for example, photolithography, Dip-Pen nanolithography, nanoimprint lithography, nanosphere lithography, nanoball lithography, nanopillar arrays, nanowire lithography, scanning probe lithography, thermochemical lithography, thermal scanning probe lithography, local oxidation nanolithography, molecular self-assembly, stencil lithography, or electron-beam lithography.
  • Functional groups in an ordered array may be located such that each functional group is less than 200 nanometers (nm), or about 200 nm, about 225 nm, about 250 nm, about 275 nm, about 300 nm, about 325 nm, about 350 nm, about 375 nm, about 400 nm, about 425 nm, about 450 nm, about 475 nm, about 500 nm, about 525 nm, about 550 nm, about 575 nm, about 600 nm, about 625 nm, about 650 nm, about 675 nm, about 700 nm, about 725 nm, about 750 nm, about 775 nm, about 800 nm, about 825 nm, about 850 nm, about 875 nm, about 900 nm, about 925 nm, about 950 nm, about 975 nm, about 1000 nm, about 1025 nm,
  • Functional groups in a random spacing may be provided at a concentration such that functional groups are on average at least about 50 nm, about 100 nm, about 150 nm, about 200 nm, about 250 nm, about 300 nm, about 350 nm, about 400 nm, about 450 nm, about 500 nm, about 550 nm, about 600 nm, about 650 nm, about 700 nm, about 750 nm, about 800 nm, about 850 nm, about 900 nm, about 950 nm, about 1000 nm, or more than 100 nm from any other functional group.
  • the substrate may be indirectly functionalized.
  • the substrate may be PEGylated and a functional group may be applied to all or a subset of the PEG molecules.
  • the substrate may be functionalized using techniques suitable for microscaled or nanoscaled structures (e.g., an ordered structure such as microwells, micropillars, single molecular arrays, nanoballs, nanopillars, or nanowires).
  • the substrate may comprise any material, including metals, glass, plastics, ceramics or combinations thereof.
  • the solid substrate can be a flow cell.
  • the flow cell can be composed of a single layer or multiple layers.
  • a flow cell can comprise a base layer (e.g., of boro silicate glass), a channel layer (e.g., of etched silicon) overlaid upon the base layer, and a cover, or top, layer.
  • a base layer e.g., of boro silicate glass
  • a channel layer e.g., of etched silicon
  • cover or top, layer.
  • the thickness of each layer can vary, but is preferably less than about 1700 pm.
  • Layers can be composed of suitable materials such as photosensitive glasses, borosilicate glass, fused silicate, PDMS, or silicon. Different layers can be composed of the same material or different materials.
  • flow cells can comprise openings for channels on the bottom of the flow cell.
  • a flow cell can comprise millions of attached target conjugation sites in locations that can be discretely visualized.
  • various flow cells of use with embodiments of the invention can comprise different numbers of channels (e.g., 1 channel, 2 or more channels, 3 or more channels, 4 or more channels, 6 or more channels, 8 or more channels, 10 or more channels, 12 or more channels, 16 or more channels, or more than 16 channels).
  • Various flow cells can comprise channels of different depths or widths, which may be different between channels within a single flow cell, or different between channels of different flow cells.
  • a single channel can also vary in depth and/or width.
  • a channel can be less than about 50 pm deep, about 50 pm deep, less than about 100 pm deep, about 100 pm deep, about 100 pm about 500 pm deep, about 500 pm deep, or more than about 500 pm deep at one or more points within the channel.
  • Channels can have any cross sectional shape, including but not limited to a circular, a semi-circular, a rectangular, a trapezoidal, a triangular, or an ovoid cross-section.
  • the proteins or peptides may be spotted, dropped, pipetted, flowed, washed or otherwise applied to the substrate.
  • a substrate that has been functionalized with a moiety such as an NHS ester
  • no modification of the protein or peptide may be required.
  • a substrate that has been functionalized with alternate moieties e.g., a sulfhydryl, amine, or linker nucleic acid
  • a crosslinking reagent e.g., disuccinimidyl suberate, NHS
  • sulphonamides may be used.
  • the proteins or peptides of the sample may be modified with complementary nucleic acid tags.
  • Photo-activatable cross linkers may be used to direct cross linking of a sample to a specific area on the substrate. Photo-activatable cross linkers may be used to allow multiplexing of protein or peptide samples by attaching each sample in a known region of the substrate.
  • Photo-activatable cross linkers may allow the specific attachment of proteins or peptides which have been successfully tagged, for example, by detecting a fluorescent tag before cross linking a protein.
  • Examples of photo-activatable cross linkers include, but are not limited to, N-5-azido-2- nitrobenzoyloxysuccinimide, sulfosuccinimidyl 6-(4'-azido-2'-nitrophenylamino)hexanoate, succinimidyl 4,4'-azipentanoate, sulfosuccinimidyl 4,4'-azipentanoate, succinimidyl 6-(4,4'- azipentanamido)hexanoate, sulfosuccinimidyl 6-(4,4'-azipentanamido)hexanoate, succinimidyl 2-((4,4'-azipentanamido)ethyl)-l,3'-dithi
  • the polypeptides may be attached to the substrate by one or more residues.
  • the polypeptides may be attached via the N terminal, C terminal, both terminals, or via an internal residue.
  • photo-cleavable linkers may be used for several different multiplexed samples.
  • photo-cleavable cross linkers may be used from one or more samples within a multiplexed reaction.
  • a multiplexed reaction may comprise control samples cross linked to the substrate via permanent crosslinkers and experimental samples cross linked to the substrate via photo-cleavable crosslinkers.
  • Each conjugated protein or peptide may be spatially separated from each other conjugated proteins or peptides such that each conjugated protein or peptide is optically resolvable. Proteins or peptides may thus be individually labeled with a unique spatial address.
  • this can be accomplished by conjugation using low concentrations of proteins or peptides, and a low density of attachment sites on the substrate so that each protein or peptide molecule is spatially separated from each other protein or peptide molecule.
  • a light pattern may be used such that proteins or peptides are affixed to predetermined locations.
  • each protein or peptide may be associated with a unique spatial address. For example, once the proteins or peptides are attached to the substrate in spatially separated locations, each protein or peptide can be assigned an indexed address, such as by coordinates. In some examples, a grid of pre-assigned unique spatial addresses may be predetermined.
  • the substrate may contain easily identifiable fixed marks such that placement of each protein or peptide can be determined relative to the fixed marks of the substrate.
  • the substrate may have grid lines and/or and“origin” or other fiducials permanently marked on the surface.
  • the surface of the substrate may be permanently or semi-permanently marked to provide a reference by which to locate cross linked proteins.
  • the shape of the patterning itself, such as the exterior border of the conjugated polypeptides, may also be used as fiducials for determining the unique location of each spot.
  • the substrate may also contain conjugated protein or peptide standards and controls.
  • Conjugated protein or peptide standards and controls may be peptides or proteins of known sequence which have been conjugated in known locations.
  • conjugated protein or peptide standards and controls may serve as internal controls in an assay.
  • the proteins or peptides may be applied to the substrate from purified protein or peptides stocks, or may be synthesized on the substrate through a process such as Nucleic Acid-Programmable Protein Array (NAPPA).
  • NAPPA Nucleic Acid-Programmable Protein Array
  • the substrate may comprise fluorescence standards. These fluorescence standards may be used to calibrate the intensity of the fluorescent signals from assay to assay. These fluorescence standards may also be used to correlate the intensity of a fluorescent signal with the number of fluorophores present in an area. Fluorescence standards may comprise some or all of the different types of fluorophores used in the assay.
  • multi- affinity reagent measurements can be performed.
  • the measurement processes described herein may utilize various affinity reagents.
  • multiple affinity reagents may be mixed together and measurements may be performed on the binding of the affinity reagent mixture to the protein-substrate conjugate.
  • measurements performed on the binding of affinity reagent mixtures may vary across different solvent conditions and/or protein folding conditions; therefore, repeated measurements may be performed on the same affinity reagent or set of affinity reagents, under such varying solvent conditions and/or protein folding conditions, in order to obtain different sets of binding measurements.
  • different sets of binding measurements may be obtained by performing repeated measurements on samples in which proteins or peptides have been enzymatically treated (e.g., with glycosidase, phosphorylase, or phosphatase) or not enzymatically treated.
  • affinity reagent generally refers to a reagent that binds proteins or peptides with reproducible specificity.
  • the affinity reagents may be antibodies, antibody fragments, aptamers, mini-protein binders, or peptides.
  • mini-protein binders may comprise protein binders that may be between 30-210 amino acids in length.
  • mini-protein binders may be designed.
  • protein binders may include peptide macrocycles, (e.g., as described in [Hosseinzadeh et al.,“Comprehensive computational design of ordered peptide macrocycles,” Science , 2017 Dec. 15; 358(6369): 1461-1466], which is incorporated herein by reference in its entirety).
  • monoclonal antibodies may be preferred.
  • antibody fragments such as Fab fragments may be preferred.
  • the affinity reagents may be commercially available affinity reagents, such as commercially available antibodies.
  • the desired affinity reagents may be selected by screening commercially available affinity reagents to identify those with useful characteristics.
  • the affinity reagents may have high, moderate, or low specificity. In some examples, the affinity reagents may recognize several different epitopes. In some examples, the affinity reagents may recognize epitopes present in two or more different proteins or peptides. In some examples, the affinity reagents may recognize epitopes present in many different proteins or peptides. In some cases, an affinity reagent used in the methods of this disclosure may be highly specific for a single epitope. In some cases, an affinity reagent used in the methods of this disclosure may be highly specific for a single epitope containing a post-translational
  • affinity reagents may have highly similar epitope specificity. In some cases, affinity reagents with highly similar epitope specificity may be designed specifically to resolve highly similar protein or peptide candidate sequences (e.g. candidates with single amino acid variants or isoforms). In some cases, affinity reagents may have highly diverse epitope specificity to maximize protein or peptide sequence coverage. In some embodiments, experiments may be performed in replicate with the same affinity probe with the expectation that the results may differ, and thus provide additional information for protein or peptide
  • affinity reagents may be designed or selected for binding specifically to one or more whole proteins, protein complexes, protein fragments, whole peptides, peptide complexes, or peptide fragments without knowledge of a specific binding epitope. Through a qualification process, the binding profile of this reagent may have been elaborated. Even though the specific binding epitope(s) are unknown, binding measurements using the affinity reagent may be used to determine protein or peptide identity. For example, a commercially-available antibody or aptamer designed for binding to a protein or peptide target may be used as an affinity reagent.
  • binding of this affinity reagent to an unknown protein or peptide may provide information about the identity of the unknown protein or peptide.
  • a collection of protein- or peptide-specific affinity reagents e.g., commercially- available antibodies or aptamers
  • the collection of protein- or peptide-specific affinity reagents may comprise about 50, 100, 200, 300, 400, 500, 600, 700, 800, 900, 1000, 2000, 3000, 4000, 5000, 10000, 20000, or more than 20000 affinity reagents. In some cases, the collection of affinity reagents may comprise all
  • affinity reagents demonstrating target-reactivity in a specific organism.
  • a collection of protein- or peptide-specific affinity reagents may be assayed in series, with binding measurements for each affinity reagent made individually.
  • subsets of the protein- or peptide-specific affinity reagents may be mixed prior to binding measurement.
  • a new mixture of affinity reagents may be selected comprising a subset of the affinity reagents selected at random from the complete set.
  • each subsequent mixture may be generated in the same random manner, with the expectation that many of the affinity reagents will be present in more than one of the mixtures.
  • protein or peptide identifications may be generated more rapidly using mixtures of protein- or peptide-specific affinity reagents.
  • such mixtures of protein- or peptide-specific affinity reagents may increase the percentage of unknown proteins or peptides for which an affinity reagent binds in any individual pass.
  • Mixtures of affinity reagents may comprise about 1%, 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, or more than 90% of all available affinity reagents.
  • Mixtures of affinity reagents assessed in a single experiment may or may not share individual affinity reagents in common. In some cases, there may be multiple different affinity reagents within a collection that bind to the same protein or peptide.
  • each affinity reagent in the collection may bind to a different protein or peptide.
  • confidence in the identity of the unknown protein or peptide being the common target of said affinity reagents may increase.
  • using multiple protein or peptide affinity reagents targeting the same protein or peptide may provide redundancy in cases where the multiple affinity reagents bind different epitopes on the same protein or peptide, and binding of only a subset of the affinity reagents targeting that protein or peptide may be interfered with by post-translational modifications or other steric hinderance of a binding epitope.
  • binding of affinity reagents for which the binding epitope is unknown may be used in conjunction with binding measurements of affinity reagents for which the binding epitope is known to generate protein or peptide identifications.
  • one or more affinity reagents may be chosen to bind amino acid motifs of a given length, such as 2, 3, 4, 5, 6, 7, 8, 9, 10, or more than 10 amino acids. In some examples, one or more affinity reagents may be chosen to bind amino acid motifs of a range of different lengths from 2 amino acids to 40 amino acids.
  • the affinity reagents may be labeled with nucleic acid barcodes.
  • nucleic acid barcodes may be used to purify affinity reagents after use.
  • nucleic acid barcodes may be used to sort the affinity reagents for repeated uses.
  • the affinity reagents may be labeled with fluorophores which may be used to sort the affinity reagents after use.
  • the family of affinity reagents may comprise one or more types of affinity reagents.
  • the methods of the present disclosure may use a family of affinity reagents comprising one or more of antibodies, antibody fragments, Fab fragments, aptamers, peptides, and proteins.
  • the affinity reagents may be modified. Examples of modifications include, but are not limited to, attachment of a detection moiety. Detection moieties may be directly or indirectly attached. For example, the detection moiety may be directly covalently attached to the affinity reagent, or may be attached through a linker, or may be attached through an affinity reaction such as complementary nucleic acid tags or a biotin streptavidin pair. Attachment methods that are able to withstand gentle washing and elution of the affinity reagent may be preferred.
  • Affinity reagents may be tagged, e.g., with identifiable tags, to allow for
  • binding events e.g., with fluorescence detection of binding events.
  • identifiable tags include: fluorophores, magnetic nanoparticles, or nucleic acid barcoded base linkers. Fluorophores used may include fluorescent proteins such as GFP, YFP, RFP, eGFP, mCherry, tdtomato, FITC, Alexa Fluor 350, Alexa Fluor 405, Alexa Fluor 488, Alexa Fluor 532, Alexa Fluor 546, Alexa Fluor 555, Alexa Fluor 568, Alexa Fluor 594, Alexa Fluor 647, Alexa Fluor 680, Alexa Fluor 750, Pacific Blue, Coumarin, BODIPY FL, Pacific Green, Oregon Green, Cy3, Cy5, Pacific Orange, TRITC, Texas Red, Phycoerythrin, and Allophcocyanin.
  • affinity reagents may be untagged, such as when binding events are directly detected, e.g., with surface plasmon resonance (S), afluorophores, magnetic
  • detection moieties include, but are not limited to, fluorophores, bioluminescent proteins, nucleic acid segments including a constant region and barcode region, or chemical tethers for linking to a nanoparticle such as a magnetic particle.
  • fluorophores for example, affinity reagents may be tagged with DNA barcodes, which can then be explicitly sequenced at their locations.
  • sets of different fluorophores may be used as detection moieties by fluorescence resonance energy transfer (FRET) detection methods.
  • FRET fluorescence resonance energy transfer
  • Detection moieties may include several different fluorophores with different patterns of excitation or emission.
  • the detection moiety may be cleavable from the affinity reagent. This can allow for a step in which the detection moieties are removed from affinity reagents that are no longer of interest to reduce signal contamination.
  • the affinity reagents are unmodified.
  • the affinity reagent is an antibody
  • the presence of the antibody may be detected by atomic force microscopy.
  • the affinity reagents may be unmodified and may be detected, for example, by having antibodies specific to one or more of the affinity reagents.
  • the affinity reagent is a mouse antibody
  • the mouse antibody may be detected by using an anti-mouse secondary antibody.
  • the affinity reagent may be an aptamer which is detected by an antibody specific for the aptamer.
  • the secondary antibody may be modified with a detection moiety as described above. In some cases, the presence of the secondary antibody may be detected by atomic force microscopy.
  • the affinity reagents may comprise the same modification, for example, a conjugated green fluorescent protein, or may comprise two or more different types of modification.
  • each affinity reagent may be conjugated to one of several different fluorescent moieties, each with a different wavelength of excitation or emission. This may allow multiplexing of the affinity reagents as several different affinity reagents may be combined and/or distinguished.
  • a first affinity reagent may be conjugated to a green fluorescent protein
  • a second affinity reagent may be conjugated to a yellow fluorescent protein
  • a third affinity reagent may be conjugated to a red fluorescent protein, thus the three affinity reagents can be multiplexed and identified by their fluorescence.
  • a first, fourth, and seventh affinity reagent may be conjugated to a green fluorescent protein
  • a second, fifth, and eighth affinity reagent may be conjugated to a yellow fluorescent protein
  • a third, sixth, and ninth affinity reagent may be conjugated to a red fluorescent protein; in this case, the first, second, and third affinity reagents may be multiplexed together while the second, fourth, and seventh affinity reagents and the third, sixth, and ninth affinity reagents form two further multiplexing reactions.
  • the number of affinity reagents which can be multiplexed together may depend on the detection moieties used to differentiate them.
  • the multiplexing of affinity reagents labeled with fluorophores may be limited by the number of unique fluorophores available.
  • the multiplexing of affinity reagents labeled with nucleic acid tags may be determined by the length of the nucleic acid bar code.
  • Nucleic acids may be deoxyribonucleic acid (DNA) or ribonucleic acid (RNA).
  • each affinity reagent can be determined prior to use in an assay.
  • the binding specificity of the affinity reagents can be determined in a control experiment using known proteins or peptides. Any appropriate experimental methods may be used to determine the specificity of the affinity reagent.
  • a substrate may be loaded with known protein or peptide standards at known locations and used to assess the specificity of a plurality of affinity reagents.
  • a substrate may contain both experimental samples and a panel of controls and standards, such that the specificity of each affinity reagent can be calculated from the binding to the controls and standards and then used to identify the
  • affinity reagents with unknown specificity may be included along with affinity reagents of known specificity, data from the known specificity affinity reagents may be used to identify proteins, and the pattern of binding of the unknown specificity affinity reagents to the identified proteins may be used to determine their binding specificity. It is also possible to reconfirm the specificity of any individual affinity reagent by using the known binding data of other affinity reagents to assess which proteins the individual affinity reagent bound. In some cases, the frequency of binding of the affinity reagent to each known protein conjugated to the substrate may be used to derive a probability of binding to any of the proteins on the substrate.
  • the frequency of binding to known proteins or peptides containing an epitope may be used to determine the probability of binding of the affinity reagent to a particular epitope.
  • an affinity reagent panel the specificities of the affinity reagents may be increasingly refined with each iteration. While affinity reagents that are uniquely specific to particular proteins may be used, methods described herein may not require them. Additionally, methods may be effective on a range of specificities. In some examples, methods described herein may be particularly efficient when affinity reagents are not specific to any particular protein or peptide, but are instead specific to amino acid motifs (e.g., the tri-peptide AAA).
  • the affinity reagents may be chosen to have high, moderate, or low binding affinities. In some cases, affinity reagents with low or moderate binding affinities may be preferred. In some cases, the affinity reagents may have dissociation constants of about 10 3 M, 10 4 M, 10 5 M, 10 6 M, 10 7 M, 10 8 M, 10 9 M, 10 10 M, or less than about 10 10 M. In some cases the affinity reagents may have dissociation constants of greater than about 10 10 M, 10 9 M, 10 8 M, 10 7 M, 10 6 M, 10 5 M, 10 4 M, 10 3 M, 10 2 M, or greater than 10 2 M. In some cases, affinity reagents with low or moderate k 0ff rates or moderate or high k on rates may be preferred.
  • affinity reagents may be chosen to bind modified amino acid sequences, such as phosphorylated or ubiquitinated amino acid sequences.
  • one or more affinity reagents may be chosen to be broadly specific for a family of epitopes that may be contained by one or more proteins.
  • one or more affinity reagents may bind two or more different proteins.
  • one or more affinity reagents may bind weakly to their target or targets. For example, affinity reagents may bind less than 10%, less than 10%, less than 15%, less than 20%, less than 25%, less than 30%, or less than 35% to their target or targets.
  • one or more affinity reagents may bind moderately or strongly to their target or targets.
  • affinity reagents may bind more than 35%, more than 40%, more than 45%, more than 60%, more than 65%, more than 70%, more than 75%, more than 80%, more than 85%, more than 90%, more than 91%, more than 92%, more than 93%, more than 94%, more than 95%, more than 96%, more than 97%, more than 98%, or more than 99% to their target or targets.
  • an excess of the affinity reagent may be applied to the substrate.
  • the affinity reagent may be applied at about a 1 : 1, 2: 1, 3 : 1, 4: 1, 5: 1, 6: 1, 7: 1, 8: 1,
  • the affinity reagent may be applied at about a 1 : 1, 2: 1, 3 : 1, 4: 1, 5: 1, 6: 1, 7: 1, 8: 1, 9: 1, or 10: 1 excess relative to the expected incidence of the epitope in the sample proteins or peptides.
  • a linker moiety may be attached to each affinity reagent and used to reversibly link bound affinity reagents to the substrate or unknown protein to which it binds.
  • a DNA tag may be attached to the end of each affinity reagent and a different DNA tag attached to the substrate or each unknown protein or peptide.
  • a linker DNA complementary to the affinity reagent-associated DNA tag on one end and the substrate- associated tag on the other may be washed over the chip to bind the affinity reagent to the substrate and prevent the affinity reagent from dissociating prior to measurement.
  • the linked affinity reagent may be released by washing in the presence of heat or high salt concentration to disrupt the DNA linker bond.
  • FIG. 21 illustrates two hybridization steps in reinforcing a binding between an affinity reagent and a protein or peptide, in accordance with some embodiments.
  • step 1 of FIG. 21 illustrates an affinity reagent hybridization.
  • an affinity reagent 2110 hybridizes to a protein or peptide 2130.
  • the protein or peptide 2130 is bound to a slide 2105.
  • the affinity reagent 2110 has a DNA tag 2120 attached.
  • an affinity reagent may have more than one DNA tag attached.
  • an affinity reagent may have 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, or more than 20 DNA tags attached.
  • DNA tag 2120 comprises a single-stranded DNA (ssDNA) tag having a recognition sequence 2125. Additionally, the protein or peptide 2130 comprises two DNA tags 2140. In some embodiments, DNA tags may be added using chemistry that reacts with cysteines in a protein. In some embodiments, a protein or peptide may have more than one DNA tag attached. In some embodiments, a protein or peptide may have 1,
  • Each DNA tag 2140 comprises an ssDNA tag having a recognition sequence 2145.
  • a DNA linker 2150 hybridizes to DNA tags 2120 and 2140 attached to the affinity reagent 2110 and to the protein or peptide 2130, respectively.
  • the DNA linker 2150 comprises ssDNA having complementary sequences to recognition sequences 2125 and 2145, respectively.
  • recognition sequences 2125 and 2145 are situated on the DNA linker 2150 so as to allow for the DNA linker 2150 to bind to both DNA tags 2120 and 2140 at the same time, as illustrated in step 2.
  • a first region 2152 of the DNA linker 2150 selectively hybridizes to recognition sequence 2125
  • a second region 2154 of the DNA linker 2150 selectively hybridizes to recognition sequence 2145.
  • the first region 2152 and the second region 2154 may be spaced apart from each other on the DNA linker 2150.
  • a first region of a DNA linker and a second region of a DNA linker may be spaced apart with a non-hybridizing spacer sequence between the first region and the second region.
  • a sequence of recognition sequence may be less than fully complementary to a DNA linker and may still bind to the DNA linker sequence.
  • a length of a recognition sequence may be less than 5 nucleotides, 5 nucleotides, 6 nucleotides, 7 nucleotides, 8 nucleotides, 9 nucleotides, 10 nucleotides, 11 nucleotides, 12 nucleotides, 13 nucleotides, 14 nucleotides, 15 nucleotides, 16 nucleotides, 17 nucleotides, 18 nucleotides, 19 nucleotides, 20 nucleotides, 21 nucleotides, 22 nucleotides, 23 nucleotides, 24 nucleotides, 25 nucleotides, 26 nucleotides, 27 nucleotides, 28 nucleotides, 29 nucleotides, or 30 nucleotides, or more than 30 nucleotides.
  • nucleotides 5 nucleotides, 6 nucleotides, 7 nucleotides, 8 nucleotides, 9 nucleotides,
  • a recognition sequence may have one or more mismatches to a complementary DNA tag sequence. In some embodiments, approximately 1 in 10 nucleotides of a recognition sequence may be mismatched with a complementary DNA tag sequence and may still hybridize with the complementary DNA tag sequence. In some embodiments, less than 1 in 10 nucleotides of a recognition sequence may be mismatched with a complementary DNA tag sequence and may still hybridize with the complementary DNA tag sequence. In some embodiments, approximately 2 in 10 nucleotides of a recognition sequence may be mismatched with a complementary DNA tag sequence and may still hybridize with the complementary DNA tag sequence. In some embodiments, more than 2 in 10 nucleotides of a recognition sequence may be mismatched with a complementary DNA tag sequence and may still hybridize with the complementary DNA tag sequence.
  • the affinity reagents may also comprise a magnetic component.
  • the magnetic component may be useful for manipulating some or all bound affinity reagents into the same imaging plane or z stack. Manipulating some or all affinity reagents into the same imaging plane may improve the quality of the imaging data and reduce noise in the system.
  • the term“detector,” as used herein, generally refers to a device that is capable of detecting a signal, including a signal indicative of the presence or absence of a binding event of an affinity reagent to a protein or peptide.
  • the signal may be a direct signal indicative of the presence or absence of a binding event, such as a surface plasmon resonance (SPR) signal.
  • SPR surface plasmon resonance
  • the signal may be an indirect signal indicative of the presence or absence of a binding event, such as a fluorescent signal.
  • a detector can include optical and/or electronic components that can detect signals.
  • the term“detector” may be used in detection methods.
  • Non-limiting examples of detection methods include optical detection, spectroscopic detection, electrostatic detection, electrochemical detection, magnetic detection, fluorescence detection, surface plasmon resonance (SPR), and the like.
  • optical detection methods include, but are not limited to, fluorimetry and UV-vis light absorbance.
  • spectroscopic detection methods include, but are not limited to, mass spectrometry, nuclear magnetic resonance (NMR) spectroscopy, and infrared spectroscopy.
  • Examples of electrostatic detection methods include, but are not limited to, gel based techniques, such as, gel electrophoresis.
  • electrochemical detection methods include, but are not limited to, electrochemical detection of amplified product after high-performance liquid chromatography separation of the amplified products.
  • proteomes are vital building blocks of cells and tissues of living organisms.
  • a given organism produces a large set of different proteins, typically referred to as the proteome.
  • the proteome may vary with time and as a function of various stages (e.g., cell cycle stages or disease states) that a cell or organism undergoes.
  • a large-scale study or measurement (e.g., experimental analysis) of proteomes may be referred to as proteomics.
  • proteomics multiple methods exist to identify proteins, including immunoassays (e.g., enzyme-linked immunosorbent assay (ELISA) and Western blot), mass spectroscopy-based methods (e.g., matrix-assisted laser desorption/ionization (MALDI) and electrospray ionization (ESI)), hybrid methods (e.g., mass spectrometric immunoassay (MSIA)), and protein microarrays.
  • immunoassays e.g., enzyme-linked immunosorbent assay (ELISA) and Western blot
  • mass spectroscopy-based methods e.g., matrix-assisted laser desorption/ionization (MALDI) and electrospray ionization (ESI)
  • MALDI matrix-assisted laser desorption/ionization
  • ESI electrospray ionization
  • hybrid methods e.g., mass spectrometric immunoassay (MSIA)
  • protein microarrays e.g.
  • Methods and systems provided herein can significantly reduce or eliminate errors in identifying proteins or peptides in a sample. Such methods and systems may achieve accurate and efficient identification of proteins and peptides within a sample of unknown proteins or peptides.
  • the protein or peptide identification may be based on calculations using information of empirical measurements of the unknown proteins or peptides in the sample. For example, empirical measurements may include binding information of affinity reagent probes which are configured to selectively bind to one or more candidate proteins or peptides, length of a protein or peptide, hydrophobicity of a protein or peptide, and isoelectric point of a protein or peptide.
  • the protein or peptide identification may be optimized to be computable within a minimal memory footprint.
  • the protein or peptide identification may comprise estimation of a confidence level that each of one or more candidate proteins or peptides is present in the sample.
  • a computer-implemented method 100 for identifying a protein or peptide within a sample of unknown proteins or peptides (e.g., as illustrated in FIG. 1).
  • the method may be applied independently to each unknown protein or peptide in the sample, to generate a collection of proteins or peptides identified in the sample.
  • Protein or peptide quantities may be calculated by counting the number of identifications for each candidate protein or peptide.
  • the method for identifying a protein or peptide may comprise receiving, by the computer, information of a plurality of empirical measurements of the unknown protein or peptide in the sample (e.g., step 105).
  • the empirical measurements may comprise (i) binding measurements of each of one or more affinity reagent probes to one or more of the unknown proteins or peptides in the sample, (ii) length of one or more of the unknown proteins or peptides; (iii) hydrophobicity of one or more of the unknown proteins or peptides; and/or (iv) isoelectric point of one or more of the unknown proteins or peptides.
  • a plurality of affinity reagent probes may comprise a pool of a plurality of individual affinity reagent probes.
  • a pool of affinity reagent probes may comprise 2, 3, 4, 5, 6, 7, 8, 9, 10, or more than 10 types of affinity reagent probes.
  • a pool of affinity reagent probes may comprise 2 types of affinity reagent probes that combined make up a majority of the composition of the affinity reagent probes in the pool of affinity reagent probes.
  • a pool of affinity reagent probes may comprise 3 types of affinity reagent probes that combined make up a majority of the composition of the affinity reagent probes in the pool of affinity reagent probes. In some embodiments, a pool of affinity reagent probes may comprise 4 types of affinity reagent probes that combined make up a majority of the composition of the affinity reagent probes in the pool of affinity reagent probes. In some embodiments, a pool of affinity reagent probes may comprise 5 types of affinity reagent probes that combined make up a majority of the composition of the affinity reagent probes in the pool of affinity reagent probes.
  • a pool of affinity reagent probes may comprise more than 5 types of affinity reagent probes that combined make up a majority of the composition of the affinity reagent probes in the pool of affinity reagent probes.
  • Each of the affinity reagent probes may be configured to selectively bind to one or more candidate proteins or peptides among the plurality of candidate proteins or peptides.
  • the affinity reagent probes may be k-mer affinity reagent probes. In some embodiments, each k-mer affinity reagent probe is configured to selectively bind to one or more candidate proteins or peptides among a plurality of candidate proteins.
  • the information of empirical measurements may comprise binding measurements of a set of probes that are believed to have bound to an unknown protein or peptide.
  • At least a portion of the information of empirical measurements of an unknown protein may be compared, by the computer, against a database comprising a plurality of protein or peptide sequences (e.g., step 110).
  • a database comprising a plurality of protein or peptide sequences.
  • Each of the protein or peptide sequences may correspond to a candidate protein or peptide among the plurality of candidate proteins or peptides.
  • the plurality of candidate proteins or peptides may comprise at least 10, at least 20, at least 30, at least 40, at least 50, at least 60, at least 70, at least 80, at least 90, at least 100, at least 150, at least 200, at least 250, at least 300, at least 350, at least 400, at least 450, at least 500, at least 600, at least 700, at least 800, at least 900, at least 1000, or more than 1000 different candidate proteins or peptides.
  • a probability that an empirical measurement on the candidate protein or peptide may generate an observed measurement outcome may be calculated or generated by the computer (e.g., in step 115).
  • the term“measurement outcome,” as used herein, refers to the information observed on performing a measurement.
  • the measurement outcome of an affinity reagent binding experiment may be a positive or negative outcome, such as either binding or non-binding of the reagent.
  • the measurement outcome of an experiment measuring the length of a protein or peptide may be 417 amino acids.
  • a probability that an empirical measurement on the candidate protein or peptide may not generate an observed measurement outcome may be calculated or generated, by the computer. Additionally, or alternatively, a probability that an empirical measurement on the candidate protein or peptide would generate an unobserved measurement outcome, may be calculated or generated by the computer. Additionally, or alternatively, a probability that a series of empirical measurements on the candidate protein or peptide may generate an outcome set may be calculated or generated, by the computer.
  • Outcome set refers to a plurality of independent measurement outcomes for a protein or peptide.
  • a series of empirical affinity reagent binding measurements may be performed on an unknown protein or peptide.
  • the binding measurement of each individual affinity reagent comprises a measurement outcome, and the set of all measurement outcomes is the outcome set.
  • the outcome set may be a subset of all observed outcomes.
  • the outcome set may consist of measurement outcomes that were not empirically observed.
  • a probability that the unknown protein or peptide is the candidate protein or peptide may be calculated or generated, by the computer.
  • steps 115 and/or 120 may be performed iteratively or non-iteratively.
  • the probabilities in step 115 may be generated based on the comparison of the empirical measurement outcomes of the unknown proteins or peptides against the database comprising the plurality of protein or peptide sequences for all candidate proteins or peptides.
  • the input to the algorithm may comprise a database of candidate protein or peptide sequences and a set of empirical measurements (e.g., probes that are believed to have bound to an unknown protein or peptide, length of the unknown protein or peptide,
  • the input to the algorithm may comprise parameters relevant to estimating the probability of any of the affinity reagents generating any binding measurement for any of the candidate proteins or peptides (e.g. trimer- level binding probabilities for each affinity reagent).
  • the output of the algorithm may comprise (i) a probability that a measurement outcome or outcome set is observed given a hypothesized candidate protein or peptide identity, (ii) the most probable identity, selected from the set of candidate proteins or peptides, for the unknown protein or peptide and the probability of that identification being correct given a measurement outcome or outcome set (e.g., in step 120), and/or (iii) a group of high-probability candidate protein or peptide identities and an associated probability that the unknown protein or peptide is one of the proteins or peptides in the group.
  • the probability that the measurement outcome is observed given that a candidate protein or peptide is the protein or peptide being measured may be expressed as:
  • peptide) is calculated completely in silico. In some embodiments, P(measurement outcome
  • peptide) may be determined empirically by acquiring the measurement in replicate experiments on an isolate of the protein or peptide candidate, and calculating the P(measurement outcome
  • peptide) is derived from a database of past measurements on the protein or peptide.
  • peptide) is calculated by generating a set of confident protein or peptide identifications from a collection of unknown proteins or peptides with the results of the measurement censored, and then calculating the frequency of the measurement outcome among the set of unknown proteins or peptides that were confidently identified as the candidate protein or peptide.
  • a collection of unknown proteins or peptides may be identified using a seed value of P(measurement outcome
  • the probability that the measurement outcome is not observed given that a candidate protein is the protein or peptide being measured may be expressed as:
  • the probability that a measurement outcome set consisting of N individual measurement outcomes is observed given that a candidate protein or peptide is the protein or peptide being measured, may be expressed as a product of the probabilities for each individual measurement outcome:
  • protein) P(measurement outcome 1
  • peptide) P(measurement outcome 1
  • the probability of the unknown protein or peptide being a candidate protein (protei ) or candidate peptide (peptidei), may be calculated based on the probability of the outcome set for each possible candidate protein or peptide.
  • the measurement outcome set comprises binding of affinity reagent probes. In some embodiments, the measurement outcome set comprises non-specific binding of affinity reagent probes.
  • the protein or peptide in the sample is truncated or degraded. In some embodiments, the protein or peptide in the sample does not contain the C-terminus of the original protein or peptide. In some embodiments, the protein or peptide in the sample does not contain the N-terminus of the original protein or peptide. In some embodiments, the protein or peptide in the sample does not contain the N-terminus and does not contain the C-terminus of the original protein or peptide.
  • the empirical measurements comprise measurements performed on mixtures of antibodies. In some embodiments, the empirical measurements comprise measurements performed on samples containing proteins from a plurality of species. In some embodiments, the empirical measurements comprise measurements performed on a sample derived from humans. In some embodiments, the empirical measurements comprise
  • the empirical measurements comprise measurements performed on samples in the presence of single amino acid variants (SAVs) caused by non-synonym ous single-nucleotide polymorphisms (SNPs).
  • the empirical measurements comprise measurements on samples in the presence of genomic structural variation, such as insertions, deletions, translocations, inversions, segmental duplications, or copy number variation (CNV) affecting the sequence of the proteins or peptides in the sample.
  • SAVs single amino acid variants
  • SNPs non-synonym ous single-nucleotide polymorphisms
  • CNV copy number variation
  • the method further comprises applying the method to all unknown proteins or peptides measured in the sample.
  • the method further comprises generating, for each of the one or more candidate proteins or peptides, a confidence level that the candidate protein or peptide matches the unknown protein or peptide being measured in the sample.
  • the confidence level may comprise a probability value.
  • the confidence level may comprise a probability value with an error.
  • the confidence level may comprise a range of probability values, optionally with a confidence (e.g., about 90%, about 95%, about 96%, about 97%, about 98%, about 99%, about 99.9%, about 99.99%, about 99.999%, about 99.9999%, about 99.99999%, about 99.999999%, about
  • the method further comprises generating a probability that a candidate protein or peptide is present in the sample.
  • the method further comprises generating protein or peptide identifications, and associated probabilities, independently for each unknown protein or peptide in the sample, and generating a list of all unique proteins or peptides identified in the sample. In some embodiments, the method further comprises counting the number of identifications generated for each unique candidate protein or peptide to determine the quantity of each candidate protein or peptide in the sample. In some embodiments, a collection of protein or peptide identifications and associated probabilities may be filtered to only contain identifications of a high score, high confidence, and/or low false discovery rate.
  • binding probabilities may be generated for affinity reagents to full-length candidate proteins or peptides. In some embodiments, binding probabilities may be generated for affinity reagents to protein or peptide fragments (e.g., a subsequence of the complete protein or peptide sequence). For example, if unknown proteins were processed and conjugated to the substrate in a manner such that only the first about 100 amino acids of each unknown protein were conjugated, binding probabilities may be generated for each protein candidate such that all binding probabilities for epitope binding beyond the first about 100 amino acids are set to zero, or alternatively to a very low probability representing an error rate.
  • a similar approach may be used if the first about 10, 20, 50, 100, 150, 200, 300, 400, or more than 400 amino acids of each protein are conjugated to the substrate.
  • a similar approach may be used if the last about 10, 20, 50, 100, 150, 200, 300, 400, or more than 400 amino acids of each protein are conjugated to the substrate.
  • binding probabilities may be generated for each peptide candidate such that all binding probabilities for epitope binding beyond the first about 6 amino acids are set to zero, or alternatively to a very low probability representing an error rate.
  • a similar approach may be used if the first about 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 20, or more than 20 amino acids of each peptide are conjugated to the substrate.
  • a similar approach may be used if the last about 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 20, or more than 20 amino acids of each peptide are conjugated to the substrate.
  • a group of potential protein or peptide candidate matches may be assigned to the unknown protein or peptide.
  • a confidence level may be assigned to the unknown protein or peptide being one of any of the protein or peptide candidates in the group.
  • the confidence level may comprise a probability value.
  • the confidence level may comprise a probability value with an error.
  • the confidence level may comprise a range of probability values, optionally with a confidence (e.g., about 90%, about 95%, about 96%, about 97%, about 98%, about 99%, about 99.9%, about 99.99%, about 99.999%, about 99.9999%, about 99.99999%, about 99.999999%, about
  • an unknown protein or peptide may match strongly with two protein or peptide candidates.
  • the two protein or peptide candidates may have high sequence similarity to each other (e.g., two protein or peptide isoforms, such as proteins or peptides with single amino acid variants compared to a canonical sequence).
  • no individual protein or peptide candidate may be assigned with high confidence, but a high confidence may be ascribed to the unknown protein or peptide matching to a single, but unknown, member of the "protein group” or“peptide group” comprising the two strongly matching protein or peptide candidates.
  • the conjugated proteins or peptides may be treated with a non specific dye and the signal from the dye measured.
  • the signal resulting from the dye may be higher than locations containing a single protein or peptides, and may be used to flag locations with multiple bound proteins or peptides.
  • the plurality of candidate proteins or peptides is generated or modified by sequencing or analyzing the DNA or RNA of the human or organism from which the sample of unknown proteins or peptides is obtained or derived.
  • the method further comprises deriving information on post- translational modifications of the unknown protein or peptide.
  • the information on post- translational modifications may comprise the presence of a post-translational modification without knowledge of the nature of the specific modification.
  • the database may be considered to be an exponential product of PTMs. For example, once a protein or peptide candidate sequence has been assigned to an unknown protein or peptide, the pattern of affinity reagent binding for the assayed protein or peptide may be compared to a database containing binding measurements for the affinity reagents to the same candidate from previous experiments. For example, a database of binding measurements may be derived from binding to a Nucleic Acid
  • NAPPA Programmable Protein Array
  • a database of binding measurements may be derived from previous experiments in which protein or peptide candidate sequences were confidently assigned to unknown proteins or peptides. Discrepancies in binding measurements between the assayed protein or peptide and the database of existing measurements may provide information on the likelihood of post-translation modification. For example, if an affinity agent has a high frequency of binding to the candidate protein or peptide in the database, but does not bind the assayed protein or peptide, there is a higher likelihood of a post-translational modification being present somewhere on the protein or peptide.
  • the location of the post translational modification may be localized to at or near the binding epitope of the affinity reagent.
  • information on specific post-translational modifications may be derived by performing repeated affinity reagent measurements before and after treatment of the
  • binding measurements may be acquired for a sequence of affinity reagents prior to treatment of the substrate with a phosphatase, and then repeated after treatment with a phosphatase.
  • Affinity reagents which bind an unknown protein or peptide prior to phosphatase treatment but not after phosphatase treatment may provide evidence of phosphorylation. If the epitope recognized by the differentially binding affinity reagent is known, the phosphorylation may be localized to at or near the binding epitope for the affinity reagent.
  • the count of a particular post-translational modification may be determined using binding measurements with an affinity reagent against a particular post- translational modification.
  • an antibody that recognizes phosphorylation events may be used as an affinity reagent.
  • the binding of this reagent may indicate the presence of at least one phosphorylation on the unknown protein or peptide.
  • the number of discrete post-translational modifications of a particular type on an unknown protein may be determined by counting the number of binding events measured for an affinity reagent specific to the particular post-translational modification.
  • a phosphorylation specific antibody may be conjugated to a fluorescent reporter.
  • the intensity of the fluorescent signal may be used to determine the number of phosphorylation-specific affinity reagents bound to an unknown protein or peptide.
  • the number of phosphorylation-specific affinity reagents bound to the unknown protein or peptide may in turn be used to determine the number of phosphorylation sites on the unknown protein or peptide.
  • evidence from affinity reagent binding experiments may be combined with pre-existing knowledge of amino acid sequence motifs or specific protein or peptide locations likely to be post-translationally modified (e.g., from dbPTM, PhosphoSitePlus, or UniProt) to derive more accurate count, identification, or localization of post-translational modification. For example, if the location of a post-translational modification is not exactly determined from affinity measurements alone, a location containing an amino acid sequence motif frequently associated with the post translational modification of interest may be favored.
  • the probabilities are iteratively generated until a
  • the predetermined condition comprises generating each of the plurality of probabilities with a confidence of at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, at least 99.9%, at least 99.99%, at least 99.999%, at least 99.9999%, at least 99.99999%, at least 99.999999%, at least 99.9999999%, at least 99.9999999%, at least 99.9999999%, at least 99.9999999%, at least
  • the method further comprises generating a paper or electronic report identifying one or more unknown proteins or peptides in the sample.
  • the paper or electronic report may further indicate, for each of the candidate proteins or peptides, a confidence level for the candidate protein or peptide being present in the sample.
  • the confidence level may comprise a probability value.
  • the confidence level may comprise a probability value with an error.
  • the confidence level may comprise a range of probability values, optionally with a confidence (e.g., about 90%, about 95%, about 96%, about 97%, about 98%, about 99%, about 99.9%, about 99.99%, about 99.999%, about 99.9999%, about 99.99999%, about 99.999999%, about 99.9999999%, about 99.99999999%, about 99.999999999%, about 99.99999999%, about 99.999999999%, about 99.999999999%, about 99.99999999%, about 99.999999999%, about 99.999999999%, about 99.999999999999%, about 99.999999999999% confidence, or above 99.9999999999999% confidence).
  • a confidence e.g., about 90%, about 95%, about 96%, about 97%, about 98%, about 99%, about 99.9%, about 99.99%, about 99.999%, about 99.9999%, about 99.99999%, about 99.999999999%, about 99.999999999%, about 99.9999999999% confidence, or above 99.99
  • the paper or electronic report may further indicate the list of protein or peptide candidates identified below an expected false discovery rate threshold (e.g., a false discovery rate below 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, 0.5%, 0.4%, 0.3%, 0.2%, or 0.1%).
  • the false discovery rate may be estimated by first sorting the protein or peptide identifications in descending order of confidence. The estimated false discovery rate at any point in the sorted list may then be calculated as 1 - avg c prob, where avg c prob is the average candidate probability for all proteins or peptides at or before (e.g., higher confidence than) the current point in the list.
  • identifications below a desired false discovery rate threshold may then be generated by returning all protein or peptide identifications before the earliest point in the sorted list where the false discovery rate is higher than the threshold.
  • identifications below a desired false discovery rate threshold may be generated by returning all proteins or peptides before, and including, the latest point in the sorted list where the false discovery rate is below or equal to the desired threshold.
  • the sample comprises a biological sample.
  • the biological sample may be obtained from a subject.
  • the method further comprises identifying a disease state or a disorder in the subject based at least on the plurality of
  • the method further comprises quantifying proteins or peptides by counting the number of identifications generated for each protein or peptide candidate.
  • the absolute quantity e.g., number of protein or peptide molecules
  • the quantity may be calculated as a percentage of the total number of unknown proteins or peptides assayed.
  • the raw identification counts may be calibrated to remove systematic error from the instrument and detection systems.
  • the quantity may be calibrated to remove biases in quantity caused by variation in detectability of protein or peptide candidates. Protein or peptide detectability may be assessed from empirical
  • the disease or disorder may be an infectious disease, an immune disorder or disease, a cancer, a genetic disease, a degenerative disease, a lifestyle disease, an injury, a rare disease or an age related disease.
  • the infectious disease may be caused by bacteria, viruses, fungi and/or parasites.
  • Non-limiting examples of cancers include Bladder cancer, Lung cancer, Brain cancer, Melanoma, Breast cancer, Non-Hodgkin lymphoma, Cervical cancer, Ovarian cancer, Colorectal cancer, Pancreatic cancer, Esophageal cancer, Prostate cancer, Kidney cancer, Skin cancer, Leukemia, Thyroid cancer, Liver cancer, and Uterine cancer.
  • Some examples of genetic diseases or disorders include, but are not limited to, multiple sclerosis (MS), cystic fibrosis, Charcot- Marie-Tooth disease, Huntington's disease, Peutz-Jeghers syndrome, Down syndrome,
  • Non-limiting examples of lifestyle diseases include obesity, diabetes, arteriosclerosis, heart disease, stroke, hypertension, liver cirrhosis, nephritis, cancer, chronic obstructive pulmonary disease (copd), hearing problems, and chronic backache.
  • injuries include, but are not limited to, abrasion, brain injuries, bruising, bums, concussions, congestive heart failure, construction injuries, dislocation, flail chest, fracture, hemothorax, herniated disc, hip pointer, hypothermia, lacerations, pinched nerve, pneumothorax, rib fracture, sciatica, spinal cord injury, tendons ligaments fascia injury, traumatic brain injury, and whiplash.
  • the method comprises identifying and quantifying small molecules (e.g. metabolites) or glycans instead of, or in addition to, proteins and peptides.
  • affinity reagents such as lectins or antibodies which bind to sugars or combinations of sugars with varying propensity, may be used to identify glycans.
  • the propensity of the affinity reagents to bind various sugars or combinations of sugars may be characterized by analyzing binding to a commercially-available glycan array.
  • unknown glycans may be conjugated to a functionalized substrate using hydroxyl -reactive chemistry and binding measurements may be acquired using the glycan-binding affinity reagents.
  • binding measurements of the affinity reagents to the unknown glycans on the substrate may be used directly to quantify the number of glycans with a particular sugar or combination of sugars.
  • one or more binding measurements may be compared to predicted binding measurements from a database of candidate glycan structures using the methods described herein to identify the structure of each unknown glycan.
  • proteins or peptides are bound to the substrate and binding measurements with glycan affinity reagents are generated to identify glycans attached to the proteins or peptides.
  • binding measurements may be made with both glycan and protein or peptide affinity reagents to generate a protein or peptide backbone sequence and conjugated glycan identifications in a single experiment.
  • metabolites may be conjugated to a functionalized substrate using chemistry targeted toward coupling groups commonly found in metabolites such as sulfhydryl, carbonyl, amine, or active hydrogen. Binding measurements may be made using affinity reagents with different propensities to particular functional groups, structural motifs, or metabolites. The resulting binding measurements may be compared to predicted binding measurements for a database of candidate small molecules, and the methods described herein may be used to identify the metabolite at each location on the substrate.
  • the empirical measurements may comprise binding of reagents that are specific for a terminus of a protein or peptide (e.g., an N-terminus or a C- terminus).
  • the empirical measurements may comprise detected signals indicative of amino acid information (e.g., detected by optical detection, spectroscopic detection, electrostatic detection, electrochemical detection, magnetic detection, fluorescence detection, surface plasmon resonance (SPR), and the like) of a protein or peptide.
  • amino acids of a protein or peptide may be selectively or directly modified (e.g., chemically modified or fluorescently labeled) such that a signal may be measured by a detector suitable to detect the modification.
  • amino acids of a protein or peptide may be selectively modified and then cleaved, such that a signal of the cleaved amino acid may be measured by a detector.
  • amino acids of a protein or peptide may be cleaved (e.g., Edman degradation), such that a signal of the cleaved amino acid may be measured by a detector.
  • amino acids of a protein or peptide may be directly identified or measured by a detector (e.g., a protein nanopore).
  • the empirical measurements may comprise differential measurements obtained following a perturbation of the proteins or peptides (e.g., a change in conditions of the proteins or peptides such as pH, and/or performing an enzymatic reaction of the proteins or peptides).
  • a perturbation of the proteins or peptides e.g., a change in conditions of the proteins or peptides such as pH, and/or performing an enzymatic reaction of the proteins or peptides.
  • Example 1 Protein and peptide identification by affinity reagent binding
  • the methods described herein may be used in combination with affinity binding reagents (e.g., aptamers or antibodies) binding measurements to analyze and/or identify proteins or peptides in a sample.
  • the measurement outcome probability to be calculated is the probability of a binding or non-binding event of an affinity binding reagent (e.g., affinity reagent or affinity probe) to a protein candidate.
  • a binding probability may be modeled as being conditional on the presence of an epitope which is recognized by the affinity binding reagent being present in the sequence of the protein or peptide.
  • an epitope may be a “trimer” (a sequence of three amino acids).
  • an affinity reagent may be designed to bind the GAV trimer, but may have off-target binding to three additional recognition sites: CLD, TYL, and LAD.
  • the binding probability can be modeled as:
  • protein) (0.25, if GAV, CLD, TYL, or IAD is present in the protein or peptide sequence; 0, otherwise ⁇ .
  • protein/peptide) (0.25, if GAV, CLD, TYL, or IAD is present in the protein or peptide sequence; 0.00001, otherwise ⁇ .
  • the probability measures the outcome of the detection of antibody binding.
  • proteins or peptides from a human-derived sample are analyzed.
  • the proteins or peptides in the sample are assumed to be represented in the human“reference” proteome (for example, as found in the Uniprot database of canonical protein sequence and functional information). That is, the protein or peptide candidate list is the set of about 21 thousand proteins and associated sequences (or modifications, variations, fragments, or a combination thereof) in the UniProt database or all peptides from each of the about 21 thousand proteins in the UniProt database.
  • a collection of unknown proteins or peptides are derived from the sample, and each unknown protein or peptide is probed in a series of affinity reagent binding experiments with the outcome (binding or no binding) measured and recorded.
  • affinity reagents or “probes,” are selected to target the most frequently observed trimers (out of about 800 possible trimers) in the protein or peptide candidate list. Outside of the targeted trimer, each probe has off-target binding to a number of additional trimers which are selected at random.
  • the probability of a probe binding to a protein or peptide sequence can be expressed as:
  • n sequence length of a protein or peptide candidate
  • q length of a recognition site (e.g., 3);
  • 5 non-specific trimer binding probability (e.g., 10 5 );
  • p specific binding probability (e.g., 0.25);
  • P(no non-specific binding) and P(no specific binding) can be expressed as:
  • FIG. 2 illustrates the sensitivity of affinity reagent probes (e.g., the percent of substrates identified with a false detection rate (FDR) of less than 1%) plotted against the number of probe recognition sites (e.g., trimer-binding epitopes) in the affinity reagent probe (ranging up to 100 probe recognition sites or trimer-binding epitopes), for three different experimental cases (with 50, 100, and 200 probes used, as denoted by the gray, black, and white circles, respectively). As seen in FIG. 2, the number of probes used has a significant effect on the ability to correctly identify proteins or peptides.
  • FDR false detection rate
  • the sensitivity is the percentage of the unknown proteins or peptides that are correctly identified with a threshold (e.g., upper limit) of less than 1% of the identifications being incorrect. For example, if each probe contains 5 recognition sites or trimer-binding epitopes (1 targeted site and 4 off-target sites), the sensitivity of protein or peptide identification is less than 10% when 50 probes are used, about 60% when 100 probes are used, and about 90% when 200 probes are used. In fact, when 300 probes are used, the sensitivity exceeds 95% (result not shown on plot).
  • This protein or peptide identification approach supports probes with many off-target binding sites. Even with 60 recognition sites or trimer-binding epitopes (1 targeted site and 59 off-target sites), identification sensitivity is about 55% in a 100-probe experiment and about 90% in a 200-probe experiment.
  • FIG. 3 illustrates the sensitivity of affinity reagent probes (e.g., the percent of substrates identified with a false detection rate (FDR) of less than 1%) plotted against the number of probe recognition sites (e.g., trimer-binding epitopes) in the affinity reagent probe (ranging up to 700 probe recognition sites or trimer-binding epitopes) for three different experimental cases (with 50, 100, and 200 probes used, as denoted by the gray, black, and white circles, respectively).
  • FDR false detection rate
  • each probe contains 100 recognition sites or trimer-binding epitopes (1 targeted site, 99 off-target sites)
  • the sensitivity of protein identification is about 1% when 50 probes are used, about 30% when 100 probes are used, and about 70% when 200 probes are used.
  • the sensitivity of protein or peptide identification is less than 1% when 50 probes are used, less than 20% when 100 probes are used, and less than 40% when 200 probes are used.
  • Example 2 Protein or peptide affinity reagent binding to proteins or peptides that have been truncated or degraded
  • FIG. 4 illustrates plots showing the sensitivity of protein identification with experiments using 100 (left), 200 (center), or 300 probes (right).
  • sensitivity of affinity reagent probes e.g., the percent of substrates identified with a false detection rate (FDR) of less than 1%) is determined for an experiment in which 4 substrates lengths are measured: (1) the intact (full) protein, (2) the 50-length N- or C-terminal fragment of the protein, (3) the 100- length N- or C-terminal fragment of the protein, and (4) the 200-length N- or C-terminal fragment of the protein.
  • N- and C-terminal fragments are denoted with solid and striped bars, respectively.
  • Each probe binds to the targeted trimer and 4 other random off-target trimers. As shown in FIG. 4, a substantial proportion of proteins (-40%) may be identified, for example, even when proteins are truncated to fragments containing only 100 amino acids and 200-probe experiments are performed.
  • FIG. 4 shows that truncated proteins containing the N-terminal fragment are slightly easier to identify (e.g., with higher sensitivity of protein identification) than fragments containing the C-terminal fragment.
  • Example 3 Protein or peptide fragments containing neither the C-terminus nor the N- terminus of the intact protein or peptide from which they are derived
  • FIG. 5 illustrates plots showing the sensitivity of protein identification with experiments using various protein fragmentation approaches. In each of the top row and the bottom row, protein identification performance is shown with 50, 100,
  • affinity reagent measurements in the 4 panels from left to right, with maximum fragment length values of 50, 100, 200, 300, 400, and 500 (as denoted by the hexagons, down pointing triangles, up-pointing triangles, diamonds, rectangles, and circles, respectively).
  • each point on each subplot represents sensitivity (protein identification rate) when using a particular fragment generation approach defined by the fragment start location and fragment length. Fragments are generated at a specific starting location on each protein indexed by distance (e.g., number of amino acids (AA) away) from the N-terminus in amino acids (as plotted on the x-axis). The end of each protein fragment is selected to generate a fragment with length 50, 100, 200, 300, 400, or 500 amino acids
  • maximum fragment length (maximum fragment length, or max fragment length values), as denoted by the hexagons, down-pointing triangles, up-pointing triangles, diamonds, rectangles, and circles, respectively. If a fragment of a given designated length cannot be generated because the protein is too short, the fragment shorter than the requested length containing the C-terminus is retained. For example, when an experiment is performed with 50 affinity reagents, only a small percentage of proteins may be identified (as plotted on the y-axis). However, when an experiment is performed with 200 affinity reagent probes using fragments with a maximum length of 200 amino acids, about 50% to about 85% of proteins may be identified (as plotted on the y-axis) depending on the fragment start site (as plotted on the x-axis).
  • the 4 subplots here show similar results as those in the top row, except that any fragments which do not match the maximum fragment length (e.g., fragments not containing the C-terminus) are discarded from analysis prior to the sensitivity and false discovery rate calculation.
  • the sensitivity of protein identification is calculated only among those proteins that may have generated a valid fragment.
  • Fragment length is a major determinant of protein
  • Example 4 Protein and peptide identification by measurement of length, hydrophobicity., and/or isoelectric point
  • the methods described herein may be applied to analyze and/or identify proteins or peptides in a sample using information from measurements on the proteins or peptides, including length, hydrophobicity, and/or isoelectric point (pi).
  • the probability of measuring a particular length for a protein or peptide query candidate can be expressed by:
  • the measurement outcome is the measured length of the unknown protein or peptide
  • the expected outcome value is the length of the protein or peptide query candidate.
  • the model also uses a coefficient of variation (CV) value which describes the expected precision of the measurement approach.
  • CV coefficient of variation
  • the probability of measuring a particular hydrophobicity for a protein or peptide is calculated using the same formula, with the expected outcome value being set to a grand average of hydropathy (gravy) score calculated from the protein or peptide candidate sequence.
  • Such a gravy score can be calculated, for example, using a Biopython tool for computational molecular biology to perform a Kyte-Doolittle computational method (e.g., as described in [Kyte et al.,“A simple method for displaying the hydropathic character of a protein,” J. Mol. Biol., 1982 May 5; 157(1): 105-32], which is incorporated herein by reference in its entirety).
  • isoelectric point (pi) is modeled with an expected pi value calculated from the protein or peptide candidate sequence using Biopython to implement the methods of Bjellqvist (e.g., as described in [Audain et al.,“Accurate estimation of isoelectric point of protein and peptide based on amino acid sequences,” Bioinformatics , 2015 November 14; 32(6):821-27], which is incorporated herein by reference in its entirety), according to the methods described in [Tabb, David L.,“An algorithm for isoelectric point estimation,”
  • FIG. 6 illustrates plots showing the sensitivity of identification of human proteins or peptides (percent of substrates identified at an FDR of less than 1%) with experiments using various combinations of types of measurements.
  • protein or peptide length, hydrophobicity, or pi measurements alone, virtually no proteins or peptides can be identified (e.g., a sensitivity ⁇ 1%). Combining all three types of measurements (len + hydro + pi) still yields virtually no identifications.
  • protein or peptide length, hydrophobicity, or pi measurements may be used to augment measurements from affinity reagent probe binding experiments. For example, proteins or peptides may be fractionated based on any of these characteristics, and each fraction conjugated to a different spatial location on the substrate.
  • affinity reagent binding measurements may be made, and the measurement of protein or peptide hydrophobicity, length, or pi may be determined by the spatial address of the protein or peptide.
  • Denatured proteins or peptides may be fractionated by molecular weight based on gel filtration (SDS-PAGE) or size exclusion chromatography. The length of proteins or peptides may be estimated from the molecular weight by dividing the weight by the average mass of an amino acid (111 Da). Proteins or peptides may be fractionated by hydrophobicity using hydrophobic interaction chromatography. Proteins or peptides may be fractionated by pi using ion exchange chromatography.
  • the methods described herein may be applied to analyze and/or identify proteins or peptides in a sample using information from experiments in which mixtures of affinity reagents are measured in each binding experiment. Consistent with disclosed embodiments, the identification of 1,000 unknown human proteins was benchmarked by acquiring binding measurements using pools of commercially-available antibodies from Santa Cruz Biotechnology, Inc. The 1,000 proteins were randomly selected from the Uniprot protein database, which comprises about 21,005 proteins. A list of monoclonal antibodies available from the Santa Cruz Biotechnology catalog with reactivity against human proteins was downloaded from an online antibody registry. The list contained 22,301 antibodies and was filtered to a list of 14,566 antibodies which matched to proteins in the Uniprot human protein database. The complete collection of antibodies modeled in the experiment comprised these 14,566 antibodies.
  • a binding probability was determined for the mixture to any of the unknown proteins. Note that, although the proteins are“unknown” in the sense that the goal is to infer their identity, the algorithm is aware of the true identity of each“unknown protein.” If the mixture contains an antibody against the unknown protein, a binding probability of 0.99 was assigned. If the mixture does not contain an antibody against the unknown protein, a binding probability of 0.0488 was assigned. In other words, the probability of a binding outcome for the mixture of antibodies was modeled as:
  • protein) (0.99, if mixture contains an antibody to the protein; 0.0488, otherwise ⁇ .
  • the value of 0.0488 represents the probability of a non-specific (off-target) binding event occurring for this mixture against the protein.
  • the non-specific binding probability for a mixture was modeled based on the expected probability of any individual antibody binding a protein other than its target, and the number of proteins in the mixture.
  • the probability of a non specific binding event for the mixture of antibodies is the probability of any single antibody in the mixture binding non-specifically. This probability is calculated based on the number of antibodies in the mixture ( n ), and the probability of non-specific binding (p) for any single antibody, and can be expressed by the equation:
  • the sequence of assessed binding events (50 total, 1 per mixture) was evaluated against each of the 21,005 protein candidates in the Uniprot database. More specifically, a probability of observing the sequence of binding events was calculated for each candidate. The probability was calculated by multiplying the probability of each individual mixture binding / non-binding event across all 50 mixtures measured. The binding probability was calculated in the same manner as described above, and the probability of non-binding is one minus the binding probability. The protein query candidate with the highest binding probability is the inferred identity for the unknown protein. A probability of the identification being correct for that individual protein was calculated as the probability of the top individual candidate divided by the summed probabilities of all candidates.
  • the methods described herein may be applied to analyze and/or identify proteins or peptides in a sample obtained from many different species. For example, results from sequence of affinity reagent binding experiments may be used to identify proteins or peptides in E. coli , Saccharomyces cerevisiae (yeast), or Elomo sapiens (humans), as denoted by the circles, triangles, and squares, respectively.
  • the protein or peptide candidate list may be generated from a species-specific sequence database, such as a reference proteome for the species downloaded from Uniprot.
  • FIG. 7 illustrates plots showing the sensitivity of protein or peptide identification with experiments using 50, 100, 200, or 300 affinity reagent probe passes against unknown proteins from either E. coli , yeast, or human (as denoted by the circles, triangles, and squares, respectively). Each probe binds to a targeted trimer, and 4 additional off-target sites with probability of 0.25. The sensitivity (percentage of unknown proteins identified at a false identification rate of less than 1%) for an experiment using 200 probes was about 90% for each of the three species tested.
  • the methods described herein may be applied to analyze and/or identify proteins or peptides in a sample in the presence of single amino acid variants (SAVs) caused by non- synonymous single-nucleotide polymorphisms (SNPs). Proteins or peptides that have the same sequence except for a handful of single amino acid variants (SAVs) may be difficult to distinguish. For example, in an experiment using a series of affinity reagent measurements, the canonical form of a protein or peptide may be nearly impossible to distinguish from its variant form, unless an affinity reagent which is highly-selective for the polymorphic region of the protein or peptide is included in the experiment.
  • SAVs single amino acid variants
  • SNPs non- synonymous single-nucleotide polymorphisms
  • L other is the summed likelihood of all protein query candidates except the canonical protein and the variant protein and is a number greater than or equal to zero.
  • groups of potential protein or peptide identifications may be returned for an unknown protein or peptide.
  • the probability for the top two most likely protein query candidates may be expressed as:
  • a confident identification may be derived from the unknown protein or peptide, albeit one that does not resolve the canonical protein or peptide and the variant protein or peptide.
  • cases where L other is near zero may be likely to result in a confident identification.
  • a probabilistic model used in one or more methods described herein may be improved iteratively using empirical measurements during the computation of protein
  • the binding probabilities for each affinity reagent probe are initialized with an estimate. For example, a collection of 200 probes may each target a single trimer and have an estimated binding probability of 0 5 Proteins and peptides are identified using the approaches disclosed elsewhere herein (for example, see Example 1). Next, the binding probabilities for each probe are refined iteratively based on empirical measurements, as summarized by the steps below:
  • FIG. 8 illustrates a plot showing the binding probability (y-axis, left) and sensitivity of protein identification (y-axis, right) against iteration (x-axis).
  • thin lines show the probe binding probabilities for each individual probe, the dark line among the thin lines is the median probe binding probability, and the thick line shows the protein identification sensitivity at each iteration.
  • Example 9 Estimating identification false discovery rate from protein or peptide candidate match probabilities
  • a probabilistic model for protein and peptide inference or identification used in one or more methods described herein yields as direct results a list of protein or peptide sequence matches for each unknown protein or peptide and an associated probability of that sequence match being correct. In many cases, only a subset of the protein or peptide identifications may be correct. Therefore, a method useful for estimating and controlling the false identification rate for a set of proteins or peptides is described below.
  • prot5 probability ps 0.8 prot6 probability (/3 ⁇ 4): 0.75
  • the expected false discovery rate at each point in the list is calculated as 1— p where p is the average of all probabilities at the given point and earlier in the list (as given below):
  • a comparison of the estimated false identification rate to the true false identification rate for a simulated 200-probe experiment demonstrates accurate false identification rate estimation.
  • identification sensitivity is compared to the true false identification rate and the estimated false identification rate.
  • the estimated false identification rate is plotted against the true false identification rate (as indicated by the solid line), while the dashed line indicates an ideal perfectly accurate false identification rate estimation.
  • the estimated false identification (ID) rate may be used to threshold a list of protein identifications depending on a tolerance for false identifications.
  • Example 10 Derivation of a false discovery rate estimation approach
  • prot 3 - RGL2 prot 3 - RGL2
  • p 3 0.92
  • prot 4 - MTLR prot 4 - MTLR
  • the expected number of false discoveries in this list is 1 - the average matching probability for all proteins or peptides in the list. In this case:
  • the expected true discovery rate (# correct IDs / # IDs) is the average candidate probability:
  • the false discovery rate is 1 - true discovery rate, or:
  • Example 11 Protein and peptide identification using binding measurement outcomes
  • the methods described herein may be applied to different subsets of data associated with the binding and/or non-binding of affinity reagents to unidentified proteins or peptides.
  • methods described herein may be applied to experiments in which a particular subset of the measured binding outcomes is not considered (e.g., non-binding measurement outcomes). These methods where a subset of the measured binding outcomes are not considered may be referred to herein as a“censored” inference approach (e.g., as described in Example 1).
  • a“censored” inference approach e.g., as described in Example 1
  • the protein or peptide identifications that result from the censored inference approach are based on assessing occurrences of binding events associated with the particular unidentified proteins or peptides. Accordingly, the censored inference approach does not consider non-binding outcomes in determining identities of unknown proteins or peptides.
  • censored inference approach is in contrast to an“uncensored” approach, in which all obtained binding outcomes are considered (e.g., both binding measurement outcomes and non-binding measurement outcomes associated with the particular unidentified proteins).
  • a censored approach may be applicable in cases where there is an expectation that particular binding measurements or binding measurement outcomes are more error-prone or likely to deviate from the expected binding measurement outcome for the protein or peptide (e.g. the probability of that binding measurement outcome being generated by the protein or peptide).
  • probabilities of binding measurement outcomes and non-binding measurement outcomes may be calculated based on binding to denatured proteins or peptides with predominantly linear structure. In these conditions, epitopes may be easily accessible to affinity reagents.
  • affinity reagent binding experiment probabilities of binding measurement outcomes and non-binding measurement outcomes may be calculated based on binding to denatured proteins or peptides with predominantly linear structure. In these conditions, epitopes may be easily accessible to affinity reagents.
  • affinity reagents in an
  • binding measurements on the assayed protein or peptide sample may be collected under non-denaturing or partially-denaturing conditions where proteins are present in a“folded” state with significant 3-dimensional structure, which can in many cases cause affinity reagent binding epitopes on the protein that are accessible in a linearized form to be inaccessible due to steric hindrance in the folded state. If, for example, the epitopes that the affinity reagent recognizes for a protein are in structurally accessible regions of the folded protein, the expectation may be that empirical binding measurements acquired on the unknown sample will be consistent with the calculated probabilities of binding derived from linearized proteins.
  • the epitopes recognized by the affinity reagent are structurally inaccessible, the expectation may be that there will be more non-binding outcomes than expected from calculated probabilities of binding derived from linearized proteins.
  • the 3-dimensional structure may be configured in a number of different possible configurations, and each of the different possible configurations may have an unique expectation for binding a particular affinity reagent based on the degree of accessibility of the desired affinity reagent.
  • non-binding outcomes may be expected to deviate from the calculated binding probabilities for each protein or peptide, and a censored inference approach which only considers binding outcomes may be appropriate.
  • a censored inference approach which only considers binding outcomes may be appropriate.
  • only measured binding outcomes are considered (in other words, either non-binding outcomes are not measured, or measured non-binding outcomes are not considered), such that the probability of a binding outcome set only considers the M measured binding outcomes that resulted in a binding measurement, which is a subset of the N total measured binding outcomes containing both binding and non-binding measurement outcomes. This may be described by the expression:
  • protein) P(binding event 1
  • peptide) P(binding event 1
  • a scaling factor may be calculated for each candidate protein or peptide by dividing the P(binding outcome set
  • L For a protein or peptide of length L, with trimer recognition sites, there may be L-2 potential binding sites (e.g., every possible length L subsequence of the complete protein or peptide sequence), such that:
  • the probability of any candidate protein or peptide selected from a collection of Q possible candidate proteins or peptides, given the outcome set, may be given by:
  • the protein or peptide identification sensitivity (e.g., percent of unique proteins or peptides identified) is plotted against the number of affinity reagent cycles measured for both censored inference and uncensored inference used on linearized protein or peptide substrates.
  • the affinity reagents used are targeted against the top most abundant trimers in the proteome, and each affinity reagent has off-target affinity to four additional random trimers.
  • the uncensored approach outperforms the censored approach by a greater than ten-fold margin when 100 affinity reagent cycles are used. The degree to which uncensored inference outperforms censored inference lessens when more cycles are used.
  • Example 12 Tolerance of protein and peptide identification to random false negative and false positive affinity reagent binding
  • “False negative” binding outcomes manifest as affinity reagent binding measurements occurring less frequently than expected. Such“false negative” outcomes may arise, for example, due to issues with the binding detection method, the binding conditions (for example, temperature, buffer composition, etc.), corruption of the protein or peptide sample, or corruption of the affinity reagent stock.
  • “false positive” binding outcomes manifest as affinity reagent binding measurements occurring more frequently than expected.
  • the tolerance to“false positive” binding outcomes was assessed by switching a subset of binding outcomes from non-binding outcomes to binding outcomes. The results of this assessment are provided in Table 3.
  • Example 13 Performance of protein or peptide inference with overestimated or underestimated affinity reagent binding probabilities
  • Protein or peptide identification sensitivity was assessed using protein or peptide identification with correctly estimated affinity reagent to trimer binding probabilities, and with overestimated or underestimated affinity reagent binding probabilities.
  • the true binding probability was 0.25.
  • the underestimated binding probabilities were: 0.05, 0.1, and 0.2.
  • the overestimated binding probabilities were 0.30, 0.50, 0.75, and 0.90.
  • 300 cycles of affinity reagent measurements were acquired. None (0), all 300, or a subset (1, 50, 100, 200) of the affinity reagents had the overestimated or underestimated binding probabilities applied. All others had the correct binding probabilities (0.25) used in protein or peptide identification.
  • Table 4 The results of the analysis are provided in Table 4.
  • affinity reagents may possess a number of binding sites (e.g., epitopes) which are unknown.
  • the sensitivity of censored protein or peptide identification and uncensored protein or peptide identification approaches with affinity reagent binding measurements were compared using affinity reagents that each bind five trimer sites (e.g. a targeted trimer, and four random off-target sites) with probability 0.25 that are input into the protein or peptide identification algorithm.
  • trimer sites e.g. a targeted trimer, and four random off-target sites
  • a subset of the affinity reagents (0 of 300, 1 of 300, 50 of 300, 100 of 300, 200 of 300, or 300 of 300) had either 1, 4, or 40 additional extra binding sites each against a random trimer with binding probability 0.05, 0.1 or 0.25.
  • Table 5 The results of the analysis are shown in Table 5
  • Example 15 Performance of protein or peptide inference approaches using affinity reagents with missing binding epitopes
  • affinity reagents with a number of annotated binding epitopes that do not exist (e.g., extra expected binding sites). That is, the model used to generate expected binding probabilities for an affinity reagent contains extra expected sites that do not exist.
  • affinity reagents that each bind random trimer sites (e.g. a targeted trimer, and four random off-target sites) with probability 0.25 that are input into the protein or peptide identification algorithm.
  • a subset of the affinity reagents (0 of 300, 1 of 300, 50 of 300, 100 of 300, 200 of 300, or 300 of 300) had either 1, 4, or 40 extra expected binding sites each against a random trimer with binding probability 0.05, 0.1 or 0.25 added to the model for the affinity reagent used by the protein or peptide inference algorithm.
  • Example 16 Censored inference for affinity reagent binding analysis with an alternative scaling strategy
  • Example 11 scales the probability of an observed outcome for a protein or peptide based on the number of potential binding sites on the protein or peptide (protein/peptide length -
  • P(trimer j ) is the frequency with which the trimer occurs relative to the summed count of all 8,000 trimers in the proteome.
  • P(trimer j ) is the frequency with which the trimer occurs relative to the summed count of all 8,000 trimers in the proteome.
  • the number of successful binding events observed for a protein or peptide of length k may follow a Poisson-Binomial distribution with n trials, where n is the number of probe binding measurements made for the protein or peptide and the parameters p probes, k °f the distribution indicate the probability of success for each trial:
  • the probability of generating N binding events from a protein or peptide of length k may be given by the probability mass function of the Poisson binomial distribution ( PMF PoiBin ) parameterized by p, evaluated at N:
  • the methods described herein may be applied to any set of affinity reagents.
  • the protein or peptide identification approach may be applied to a set of affinity reagents targeting the most abundant trimers in the proteome, or targeting random trimers.
  • the results from a human protein inference analysis using affinity reagents targeting the top 300 least abundant trimers in the proteome, 300 randomly selected trimers in the proteome, or the 300 most abundant trimers in the proteome, are shown in Tables 7A-7C, respectively.
  • each affinity reagent had a binding probability of 0.25 to the targeted trimer, and a binding probability of 0.25 to 4 additional randomly selected trimers.
  • the performance of each affinity reagent set is measured based on sensitivity (e.g., the percentage of proteins identified).
  • Each affinity reagent set was assessed in 5 replicates, with the performance of each replicate plotted as a dot, and a vertical line connecting replicate measurements from the same set of affinity reagents.
  • the results from the affinity reagent set consisting of the top 300 most abundant affinity reagents is in blue, the bottom 300 in green.
  • a total of 100 different sets of 300 affinity reagents targeting random trimers were generated and assessed. Each of those sets is represented by a set of 5 grey points (one for each replicate) connected by a vertical grey line. According to the uncensored inference used in this analysis, targeting more abundant trimers improves identification performance as compared to targeting random trimers.
  • Example 18 Affinity reagents with biosimilar off-target sites
  • affinity reagent binding experiment with affinity reagents having different types of off-target binding sites (epitopes).
  • performance with two classes of affinity reagents are compared: random, and “biosimilar” affinity reagents.
  • the results from these assessments are shown in Tables 8A-8D.
  • the biosimilar affinity reagents have off-target binding sites that are biochemically similar to the targeted epitope. Both the random and biosimilar affinity reagents recognize their target epitope (e.g., a trimer) with binding probability 0.25. Each of the random class of affinity reagents has 4 randomly selected off-target trimer binding sites with binding probability 0.25. In contrast, the 4 off-target binding sites for the “biosimilar” affinity reagents are the four trimers most similar to the trimer targeted by the affinity reagent, which are bound with probability 0.25. For these biosimilar affinity reagents, the similarity between trimer sequences is computed by summing the BLOSUM62 coefficient for the amino acid pair at each sequence location.
  • Both the random and biosimilar affinity reagent sets target the top 300 most abundant trimers in the human proteome, where abundance is measured as the number of unique proteins containing one or more instances of the trimer.
  • FIG. 17 shows the performance of the censored (dashed lines) and uncensored (solid lines) protein inference approaches in terms of the percent of proteins identified in a human sample when affinity reagents with random (blue) or biosimilar (orange) off-target sites are used.
  • uncensored inference outperforms censored inference, with uncensored inference performing better in the case of biosimilar affinity reagents, and censored inference performing better in the case of random affinity reagents.
  • an optimal set of trimer targets may be chosen for a particular approach based on the candidate proteins or peptides that may be measured (for example, the human proteome), the type of protein or peptide inference being performed (censored or uncensored), and the type of affinity reagents being used (random or biosimilar).
  • A“greedy” algorithm as described below, may be used to select a set of optimal affinity reagents:
  • i Simulate binding of the candidate AR against the protein or peptide set.
  • ii. Perform protein or peptide inference for each protein or peptide using the simulated binding measurements from the candidate AR and the simulated binding measurements from all previously selected ARs.
  • affinity reagents selected by the greedy optimization algorithm improves the performance of both random and biosimilar affinity reagent sets using both censored protein or peptide inference and uncensored protein or peptide inference approaches. Additionally, random affinity reagents sets perform almost identically to biosimilar affinity reagents sets when the greedy approach is used to select affinity reagents.
  • Example 19 Protein or peptide inference using binding of mixtures of affinity reagents
  • the methods described herein may be applied to analyze and/or identify proteins or peptides that have been measured using mixtures of affinity reagents.
  • the probability of a specific protein or peptide generating a binding outcome when assayed by a mixture of affinity reagents may be computed as follows: 1) Calculate p ⁇ s ⁇ , the average probability of non-specific epitope binding of each affinity reagent in the mixture.
  • each affinity reagent binds 20 additional off-target sites with binding probability scaled depending on the sequence similarity between the off-target site and the targeted trimer calculated using the BLOSUM62 substitution matrix.
  • the probability for these additional off target sites is: 0.25 * 1.5 s ° T ⁇ ssei f w here
  • S 0T is the BLOSUM62 similarity between the off-target site and the targeted site
  • S se is the BLOSUM62 similarity between the targeted sequence and itself.
  • Any off-target sites with binding probability below 2.45 x 10 8 are adjusted to have binding probability 2.45 x 10 8 .
  • the non-specific epitope binding probability is 2.45 x 10 8 in this example.
  • An optimal set of 300 mixtures of affinity reagents were generated for both censored and uncensored protein or peptide inference using a greedy approach:
  • FIG. 19 shows the protein or peptide identification sensitivity when the unmixed candidate affinity reagents are used with censored protein or peptide inference and uncensored protein or peptide inference, and when mixtures are used.
  • the data plotted in FIG. 19 is shown in Tables 10A-10B. Tables 10A-B
  • Example 20 Glycan identification with a database of 7 candidate glvcans
  • this information arrives incrementally, and therefore may be computed iteratively.
  • the initial, un-normalized probability of a glycan is calculated as the product of the probabilities for each candidate glycan:
  • the size normalization is computed, which refers to the number of ways some number of affinity reagents may land on a given glycan, as a function of the number of potential binding sites of the glycan.
  • the size normalization is given by the Choose(sites_i, n) term. For example, candidate ID 52 has 6 disaccharide sites and a size normalization of [6 choose 4] which is 15. If there are more binding events than the number of available disaccharide sites, the size normalization factor is set to 1.
  • the un-normalized probabilities of each glycan are normalized to take into account this size correction by dividing by the size normalization which gives:
  • the probabilities are normalized such that the entire set of probabilities over the entire database sums up to one. This is achieved by summing the size-normalized probabilities to 0.00390641 and dividing each of the size-normalized probabilities by this normalization to achieve the final balanced probabilities:
  • Example 21 Performance of censored protein or peptide identification in samples containing protein isoforms
  • the protein and peptide identification approaches described herein may be applied to samples containing protein isoforms or peptides derived from protein isoforms.
  • An isoform of a canonical protein may refer to a variant of the canonical protein formed by alternative splicing of the same gene as the canonical protein or another gene in the same gene family as the canonical protein.
  • a protein isoform may be structurally similar to the canonical protein, typically sharing large portions of sequence with the canonical protein.
  • Peptides derived from protein isoforms may have the same sequence as a peptide from the canonical protein if, for example, the peptide is generated from a region of the protein sequence that is the same in both the canonical protein and the isoform protein.
  • Peptides derived from protein isoforms may have a different sequence from that of the canonical protein if the peptide is generated from a region of the protein sequence that is different in the canonical protein and the isoform protein.
  • a peptide spanning a splice junction may have a sequence unique to a specific protein isoform.
  • an affinity reagent binding analysis was performed on a collection of proteins consisting of 20,374 unique canonical human proteins and 21,987 unique isoforms of those canonical proteins.
  • the canonical proteins and isoform proteins are those listed in the reference human proteome available as part of the Uniprot database. Only proteins with the“Swiss-Prot” designation, used to designate proteins that have been manually annotated and reviewed, were included in the analysis.
  • the number of isoforms included for each individual canonical protein ranged from 0 to 36 isoforms.
  • the mean number of isoforms for a canonical protein in this set is 1.08.
  • the sample was analyzed using 384 affinity reagent cycles, each cycle measuring binding outcomes of a unique affinity reagent to each of the proteins in the sample.
  • Each affinity reagent binds a targeted trimer with a probability of 0.25, and to the four trimers most similar to the targeted trimer with a probability of 0.25.
  • Other off-target trimers are bound with a probability of the greater of the quantities 2.45xl0 8 and 0.25 * 1.5 x where x is the similarity of the off-target trimer to the trimer target subtracted from the similarity of the targeted trimer to itself.
  • the similarity between trimer sequences can be computed by, for example, summing the
  • BLOSUM62 coefficient for the amino acid pair at each of the three sequence locations were selected using a greedy approach, as described in Example 18, to optimize against the human proteome.
  • Censored protein inference was performed on the binding outcomes from the sample using a database containing only the sequences for the 20,374 canonical proteins in the protein sample. Because the database used for protein inference is missing the sequences of the 21,987 protein isoforms in the sample, the results of this analysis indicate performance when the sequences of potential protein isoforms in a sample are not known. With protein inference performed in this manner, the correct protein family is identified for 83.9% of the proteins in the sample with a false discovery rate of 1%.
  • the term“protein family,” as used herein, generally refers to a set of sequences including a canonical protein sequence and all isoforms of that canonical protein sequence. The correct protein family for a protein is identified if the inferred protein identity is within the same protein family as the protein being analyzed.
  • Protein identification performance using protein families defined a priori When the grouping of canonical protein sequences and isoform protein sequences into protein families is known a priori , the identification rate for protein families may be improved by calculating protein family probabilities directly. For an individual protein being measured, the probability of the protein being a member of the protein family may be calculated by summing each of the probabilities of the individual protein sequences comprising the family. The protein family with the highest probability for the protein being analyzed is assigned as the protein family identification. When protein family probabilities are calculated in this manner, the correct protein family is identified for 97.2% of the proteins in the sample at 1% false discovery rate. In comparison, the correct protein family is identified for 89.8% of the proteins in the sample at 1% false discovery rate, when the protein family probabilities are not directly calculated.
  • Example 22 Performance of censored protein or peptide identification in samples containing proteins with single amino acid variants (SAVs)
  • a single amino acid variant (SAV) of a canonical protein or peptide generally refers to a variant of the canonical protein or peptide which differs by a single amino acid.
  • Single amino acid variant proteins or peptide may typically arise from missense single nucleotide polymorphisms (SNPs) in the gene encoding the protein or peptide.
  • an affinity reagent binding analysis was performed on a collection of proteins consisting of 20,374 unique canonical human proteins and 12,827 unique SAVs of those canonical proteins.
  • the canonical proteins are those listed in the reference human proteome available as part of the Uniprot database.
  • For each canonical protein if one or more SAVs for the protein exist in the SAV database, a randomly chosen SAV is included in the sample.
  • the SAV database used is the Uniprot human polymorphisms and disease mutations index. Only proteins with the“Swiss- Prot” designation, used to designate proteins that have been manually annotated and reviewed, were included in the analysis.
  • the sample was analyzed using 384 affinity reagent cycles, each cycle measuring binding outcomes of a unique affinity reagent to each of the proteins in the sample.
  • Each affinity reagent binds a targeted trimer with a probability of 0.25, and to the four trimers most similar to the targeted trimer with a probability of 0.25.
  • Other off-target trimers are bound with a probability of the greater of the quantities 2.45xl0 8 and 0.25 * 1.5 x where x is the similarity of the off-target trimer to the trimer target subtracted from the similarity of the targeted turner to itself.
  • the similarity between trimer sequences may be computed by, for example, summing the BLOSUM62 coefficient for the amino acid pair at each of the three sequence locations.
  • Affinity reagent trimer targets were selected using a greedy approach, as described in Example 18, to optimize against the human proteome.
  • Censored protein inference was performed on the binding outcomes from the sample using a database containing only the sequences for the 20,374 canonical proteins in the protein sample. Because the database used for protein inference is missing the sequences of the 12,827 SAV proteins in the sample, the results of this analysis indicate performance when the sequences of all potential SAVs in a sample are not known. With protein inference performed in this manner, the correct SAV protein family is identified for 96.0% of the proteins in the sample with a false discovery rate of 1%.
  • the term“SAV protein family,” as used herein, generally refers to set of sequences including a canonical protein sequence and all SAVs of that canonical protein sequence. The correct SAV protein family for a protein is identified if the inferred protein identity is within the same SAV protein family as the protein being analyzed.
  • the identification rate for SAV protein families may be improved by calculating SAV protein family probabilities directly.
  • the probability of the protein being a member of a SAV protein family may be calculated by summing each of the probabilities of the individual protein sequences comprising the family.
  • the SAV protein family with the highest probability for the protein being analyzed is assigned as the SAV protein family identification.
  • SAV protein family probabilities are calculated in this manner, the correct SAV protein family is identified for 96.5% of the proteins in the sample at 1% false discovery rate. In comparison, the correct SAV protein family is identified for 96.1% of the proteins in the sample at 1% false discovery rate when the protein family probabilities are not directly calculated.
  • Example 23 Performance of censored protein or peptide inference on a sample containing proteins or peptides from a mixture of species
  • a protein or peptide sample may comprise proteins or peptides from each of a plurality of species.
  • a protein or peptide sample may contain proteins or peptides arising from external sources such as fossils.
  • a protein or peptide sample may contain proteins or peptides that are synthesized, modified, or engineered, such as a recombinant protein, or a protein or peptide synthesized by in-vitro transcription and translation.
  • synthesized, modified, or engineered proteins or peptides may contain non-natural sequences (e.g., arising from CRISPR-Cas9 modification or other artificial gene constructs) or modifications, variations, or fragments thereof.
  • Each of the species may be, for example, an animal such as a mammal (e.g., human, mouse, rat, primate, or simian), farm animals (production cattle, dairy cattle, poultry, horses, pigs, and the like), sport animals, companion animals (e.g., pet or support animals); a plant, a protist, a bacterium, a virus, or an archeon.
  • a mammal e.g., human, mouse, rat, primate, or simian
  • farm animals production cattle, dairy cattle, poultry, horses, pigs, and the like
  • sport animals e.g., pet or support animals
  • a plant e.g., a protist, a bacterium, a virus, or an archeon.
  • a sample from a mouse tumor xenograft model may comprise substantial amounts of proteins or peptides of both mouse and human origin.
  • an affinity reagent binding analysis was performed on a collection of proteins consisting of 2,000 unique mouse proteins and 2,000 unique human proteins. Both the human proteins and the mouse proteins were randomly selected from the collection of canonical Swiss-Prot sequence entries in the Uniprot reference proteome of the respective species. The sample was analyzed using 384 affinity reagent cycles, each cycle measuring binding outcomes of a unique affinity reagent to each of the proteins in the sample.
  • Each affinity reagent binds a targeted trimer with a probability of 0.25, and to the four trimers most similar to the targeted trimer with a probability of 0.25.
  • Other off-target trimers are bound with probability the greater of the quantities 2.45xl0 8 and 0.25 * 1.5 x where x is the similarity of the off-target trimer to the trimer target subtracted from the similarity of the targeted trimer to itself.
  • the similarity between trimer sequences may be computed by, for example, summing the BLOSUM62 coefficient for the amino acid pair at each of the three sequence locations.
  • Affinity reagent trimer targets were selected using a greedy approach, as described in Example 18, to optimize against the human proteome.
  • Example 24 Design of an affinity reagent set against a targeted panel of proteins or peptides
  • a set of affinity reagents may be designed that is optimized for identification of a specific subset of proteins or peptides in a sample. For example, an optimal collection of affinity reagents can be used to identify a specific set of target proteins in fewer affinity reagent binding cycles as compared to using a set optimized for identification of the entire proteome.
  • a set of affinity reagents is generated for optimal identification of 25 human proteins, which are potential biomarkers for clinical response to cancer immunotherapy treatment. The proteins in the targeted panel are listed in Table 11.
  • Example 18 To generate a set of affinity reagents optimized for identification of the complete proteome, a greedy selection approach, as described in Example 18, was applied. This set of affinity reagents can be referred to as the“proteome-optimized” affinity reagent set.
  • a modified version of step 4) i) in Example 18 is performed, in which, rather than calculating the score for the candidate affinity reagent by summing each of the probabilities of the correct protein identification for each protein determined by protein inference, the score for the candidate affinity reagent is calculated by summing each of the probabilities of the correct protein identification for only the proteins in the targeted panel.
  • This affinity reagent set can be referred to as the“panel-optimized” affinity reagent set.
  • the performance of the proteome-optimized and panel-optimized affinity reagent sets were tested on a human proteome sample containing every unique, canonical protein in the Swiss-Prot human reference proteome from Uniprot (20,374 proteins). This sample includes all 25 of the proteins in the target panel. Both affinity reagents sets were used to analyze the protein sample, and censored inference used to generate protein identifications for every protein in the sample.
  • the number of targeted panel proteins identified by the proteome-optimized and panel-optimized affinity reagent sets is indicated in Table 12. For a targeted panel protein to be counted as a successful identification, it must be present in the list of all proteins identified in the sample at a false discovery rate below 1%. Identification was performed with varying number of affinity reagent cycles. For example, 150 affinity reagent cycles indicates that protein inference was performed on a dataset comprising analysis with the first 150 affinity reagents from either the proteome-optimized or panel-optimized set, with each affinity reagent analyzed in an individual cycle.
  • Example 25 Performance of protein or peptide inference using incomplete measurements [such as detection of presence, count or order of individual amino acids]
  • the protein or peptide inference approach described herein may be applied to incomplete measurements of the protein or peptide, such as measurements of a subset of the amino acids in a sample of unknown proteins and peptides (e.g., only specific amino acids in proteins and peptides).
  • an unknown protein or peptide in a sample can be identified or inferred based on detecting, measuring, or identifying only a subset of the amino acids in the protein or peptide.
  • amino acid characteristics e.g., presence, count, or order
  • measurements on a protein or peptide may be made which indicate the presence or absence of an amino acid in a protein or peptide (binary), the count of an amino acid in a protein or peptide (count), and/or the order of amino acids in a protein or peptide (order).
  • the empirical measurements may comprise detected signals indicative of amino acid information (e.g., detected by optical detection, spectroscopic detection, electrostatic detection, electrochemical detection, magnetic detection, fluorescence detection, surface plasmon resonance (SPR), and the like) of a protein or peptide.
  • amino acids of a protein or peptide may be selectively modified (e.g., chemically modified or fluorescently labeled) such that a signal may be measured by a detector suitable to detect the modification.
  • amino acids of a protein or peptide may be directly identified or measured by a detector (e.g., a protein nanopore).
  • the empirical measurements may comprise differential measurements obtained following a perturbation of the proteins or peptides (e.g., a change in conditions of the proteins or peptides such as pH, and/or performing an enzymatic reaction of the proteins or peptides).
  • a perturbation of the proteins or peptides e.g., a change in conditions of the proteins or peptides such as pH, and/or performing an enzymatic reaction of the proteins or peptides.
  • proteins or peptides may be modified by a series of reactions which each selectively modify a particular amino acid.
  • Each reaction of the series of reactions may have a reaction efficiency between 0 and 1, indicating the probability of the reaction successfully modifying any single amino acid substrate within the protein or peptide.
  • the presence or absence of a selectively-modified amino acid may be detected, the count of a selectively-modified amino acid may be detected, and/or the order of a particular set of selectively-modified amino acids within the protein or peptide may be detected.
  • a subset of amino acids of an unknown protein or peptide may be detected (e.g., my measuring conductance through a pore of a nanopore sequencer). For example, the presence or absence, the count, and/or the order of a set of certain amino acids within the protein or peptide may be detected.
  • an unknown protein is subjected to trypsin digestion to produce a plurality of peptides, each of which can be subjected to peptide inference as described herein. After each of the tryptic peptides is identified, protein inference can be performed based on the identified constituent peptides to identify the unknown protein.
  • protein/peptide) can be expressed as 1 - (1 - R aa ) Caa where R aa is the reaction efficiency for the amino acid and Caa is the count of the number of times the amino acid occurs in the protein or peptide.
  • protein/peptide) can be expressed as 1 - Pr(amino acid detected present
  • the probability of a particular candidate protein or peptide being the correct identification for the protein or peptide being measured can be expressed, for example, as
  • a count of a subset of amino acids (e.g., a set of selectively- modified amino acids) in a protein or peptide may be measured, and protein or peptide identifications may be generated based on such count measurements (e.g., with or without knowledge of the absolute or relative locations of such amino acids).
  • amino acids of a protein or peptide may be selectively modified (e.g., chemically modified or fluorescently labeled) such that a signal of the amino acids may be measured by a detector suitable to detect the modification.
  • amino acids of a protein or peptide may be directly identified or measured by a detector (e.g., by a protein nanopore or Edman degradation) to determine the count of the subset of amino acids in a protein or peptide.
  • protein/peptide) can be expressed as ( b aa ) M * (1— R a a) Caa ⁇ M * Raais the reaction efficiency for the amino acid, Caa is the count of the number of times the amino acid occurs in the protein or peptide, and M is the measured count for the amino acid in the protein or peptide. If M > Caa, a probability of 0 is returned. If a sequence of multiple amino acid count
  • the probabilities may be multiplied to determine the probability of the complete set of N measurements given a candidate protein or peptide, as expressed by:
  • an order of a subset of amino acids (e.g., a set of selectively- modified amino acids) in a protein or peptide may be measured, and protein or peptide identifications may be generated based on such order measurements (e.g., with or without knowledge of the absolute or relative locations of such amino acids).
  • amino acids of a protein or peptide may be selectively modified (e.g., chemically modified or fluorescently labeled) such that a signal of the amino acids may be sequentially measured by a detector suitable to detect the modification.
  • amino acids of a protein or peptide may be directly identified or measured by a detector (e.g., by a protein nanopore or Edman degradation) in a sequential fashion to determine the order of the subset of amino acids in a protein or peptide.
  • a detector e.g., by a protein nanopore or Edman degradation
  • a protein or peptide with sequence“TINYPRTEIN” may generate an order measurement outcome of“ININ” if only the amino acids I and N are selectively modified and measured.
  • the same protein or peptide may generate a measurement outcome “INN,” or“UN,” in cases where a subset of the amino acid modifications and/or measurements is not successful.
  • protein/peptide) can be expressed as Pr(aa_counts
  • protein/peptide) is
  • N was measured 2 times
  • C aai is the number of times amino acid i occurs in the sequence of the candidate protein or peptide
  • amino acids 1 to L are all unique amino acids measured in the protein or peptide (e.g., I and N, for measurement outcome“ININ”). If the number of counts measured for any particular amino acid is greater than the number of times that amino acid occurs in the protein or peptide candidate sequence, then the probability Pr(aa_counts
  • NUMORDER is the number of ways a particular outcome can be generated from the protein or peptide sequence. For example, the measurement outcome of“IN” can be generated from the protein or peptide“TINYPRTEIN” in the following ways:
  • NUMORDER (“TINYPRTEIN,”“TINYPRTEIN,” and“TINYPRTEIN” ⁇ , so NUMORDER is 3 for this particular outcome and protein sequence. Note that NUMORDER has a value of zero in cases where it is not possible to generate a particular outcome from a protein or peptide (for example, the measurement outcome of“INNI” cannot be generated from the protein or peptide
  • TERTPRTEIN The probability of a particular candidate protein or peptide being the correct identification for the protein or peptide being measured can be expressed, for example, as
  • the height of each bar indicates the percent of proteins in the sample identified with a false discovery rate below 1%.
  • the sample measured was a human protein sample containing 1,000 proteins. The results indicate that a substantial number of proteins can be identified using measurements of order of amino acids with a reaction efficiency of 0.9 or higher. If measurements of counts of amino acids are used, a substantial number of proteins can be identified with a reaction efficiency of 0.99 or higher.
  • the collection of reagents for selective modification and detection of amino acids was expanded to include the 20 amino acids R, H, K, D, E, S, T, N, Q, C, G, P, A, V, I, L, M, F, Y, and W.
  • the detection modality is indicated by the line shade, and the reaction efficiency is indicated on the x-axis.
  • the y-axis indicates the percent of proteins identified with a false discovery rate below 1% in the sample.
  • FIG. 24 illustrates the performance of protein identification using measurements of order of amino acids, where amino acids are measured with a detection probability (equal to reaction efficiency) indicated on the x-axis.
  • the y-axis indicates the percent of proteins in the sample identified with a false discovery rate below 1%.
  • the experiment was performed with measurements of order of amino acids measured at the N-terminal 25, 50, 100, or 200 amino acids of each protein, and the candidate protein sequence database consisted of the first 25, 50, 100, or 200 amino acids, respectively, of each canonical protein sequence in the Uniprot reference human protein database.
  • FIG. 25 illustrates the performance of various approaches on a tryptic digest of a sample consisting of 1,000 unique human proteins.
  • the sample contains all fully tryptic peptides of length greater than 12 with no missed cleavages arising from these proteins.
  • the dark lines indicate performance when protein identification is performed using measurements of the order of all amino acids, which are measured at varying detection probability (equivalent to reaction efficiency).
  • the light lines indicate performance when only the order of amino acids K, D, W, and C are measured at varying detection probability (equivalent to reaction efficiency).
  • the sequence database used for inference contains the sequences of every fully tryptic peptide with length greater than 12 with no missed cleavages arising from these proteins, derived from every canonical protein sequence in the human reference proteome database downloaded from Uniprot.
  • the solid lines indicate the percentage of peptides in the sample identified at a false discovery rate below 1%.
  • the dashed lines indicate the percentage of proteins in the sample identified at a false discovery rate below 1%.
  • a protein is identified if a peptide with sequence unique to that protein is identified at a false discovery rate below 1%.
  • FIG. 10 shows a computer system 1001 that is programmed or otherwise configured to: receive information of empirical measurements of unknown peptides in a sample, compare information of empirical measurements against a database comprising a plurality of peptide sequences corresponding to candidate peptides, generate probabilities of a candidate peptide generating the observed measurement outcome set, and/or generate
  • the computer system 1001 can regulate various aspects of methods and systems of the present disclosure, such as, for example, receiving information of empirical measurements of unknown peptides in a sample, comparing information of empirical measurements against a database comprising a plurality of peptide sequences corresponding to candidate peptides, generating probabilities of a candidate peptide generating the observed measurement outcome set, and/or generating probabilities that candidate peptides are correctly identified in the sample.
  • the computer system 1001 can be an electronic device of a user or a computer system that is remotely located with respect to the electronic device.
  • the electronic device can be a mobile electronic device.
  • the computer system 1001 includes a central processing unit (CPU, also“processor” and“computer processor” herein) 1005, which can be a single core or multi core processor, or a plurality of processors for parallel processing.
  • the computer system 1001 also includes memory or memory location 1010 (e.g., random-access memory, read-only memory, flash memory), electronic storage unit 1015 (e.g., hard disk), communication interface 1020 (e.g., network adapter) for communicating with one or more other systems, and peripheral devices 1025, such as cache, other memory, data storage and/or electronic display adapters.
  • the memory 1010, storage unit 1015, interface 1020 and peripheral devices 1025 are in
  • the storage unit 1015 can be a data storage unit (or data repository) for storing data.
  • the computer system 1001 can be operatively coupled to a computer network (“network”) 1030 with the aid of the communication interface 1020.
  • the network 1030 can be the Internet, an internet and/or extranet, or an intranet and/or extranet that is in communication with the Internet.
  • the network 1030 in some cases is a telecommunication and/or data network.
  • the network 1030 can include one or more computer servers, which can enable distributed computing, such as cloud computing.
  • one or more computer servers may enable cloud computing over the network 1030 (“the cloud”) to perform various aspects of analysis, calculation, and generation of the present disclosure, such as, for example, receiving information of empirical measurements of unknown peptides in a sample, comparing information of empirical measurements against a database comprising a plurality of peptide sequences corresponding to candidate peptides, generating probabilities of a candidate peptide generating the observed measurement outcome set, and/or generating probabilities that candidate peptides are correctly identified in the sample.
  • cloud computing may be provided by cloud computing platforms such as, for example, Amazon Web Services (AWS), Microsoft Azure, Google Cloud Platform, and IBM cloud.
  • the network 1030 in some cases with the aid of the computer system 1001, can implement a peer-to-peer network, which may enable devices coupled to the computer system 1001 to behave as a client or a server.
  • the CPU 1005 can execute a sequence of machine-readable instructions, which can be embodied in a program or software.
  • the instructions may be stored in a memory location, such as the memory 1010.
  • the instructions can be directed to the CPU 1005, which can subsequently program or otherwise configure the CPU 1005 to implement methods of the present disclosure. Examples of operations performed by the CPU 1005 can include fetch, decode, execute, and writeback.
  • the CPU 1005 can be part of a circuit, such as an integrated circuit.
  • a circuit such as an integrated circuit.
  • One or more other components of the system 1001 can be included in the circuit.
  • the circuit is an application specific integrated circuit (ASIC).
  • ASIC application specific integrated circuit
  • the storage unit 1015 can store files, such as drivers, libraries and saved programs.
  • the storage unit 1015 can store user data, e.g., user preferences and user programs.
  • the computer system 1001 in some cases can include one or more additional data storage units that are external to the computer system 1001, such as located on a remote server that is in communication with the computer system 1001 through an intranet or the Internet.
  • the computer system 1001 can communicate with one or more remote computer systems through the network 1030.
  • the computer system 1001 can communicate with a remote computer system of a user.
  • remote computer systems include personal computers (e.g., portable PC), slate or tablet PC’s (e.g., Apple® iPad, Samsung® Galaxy Tab), telephones, Smart phones (e.g., Apple® iPhone, Android-enabled device,
  • Blackberry® or personal digital assistants.
  • the user can access the computer system 1001 via the network 1030.
  • Methods as described herein can be implemented by way of machine (e.g., computer processor) executable code stored on an electronic storage location of the computer system 1001, such as, for example, on the memory 1010 or electronic storage unit 1015.
  • the machine executable or machine readable code can be provided in the form of software.
  • the code can be executed by the processor 1005.
  • the code can be retrieved from the storage unit 1015 and stored on the memory 1010 for ready access by the processor 1005.
  • the electronic storage unit 1015 can be precluded, and machine-executable instructions are stored on memory 1010.
  • the code can be pre-compiled and configured for use with a machine having a processer adapted to execute the code, or can be compiled during runtime.
  • the code can be supplied in a programming language that can be selected to enable the code to execute in a pre compiled or as-compiled fashion.
  • Aspects of the systems and methods provided herein, such as the computer system 1001 can be embodied in programming.
  • Various aspects of the technology may be thought of as “products” or“articles of manufacture” typically in the form of machine (or processor) executable code and/or associated data that is carried on or embodied in a type of machine readable medium.
  • Machine-executable code can be stored on an electronic storage unit, such as memory (e.g., read-only memory, random-access memory, flash memory) or a hard disk.
  • “Storage” type media can include any or all of the tangible memory of the computers, processors or the like, or associated modules thereof, such as various semiconductor memories, tape drives, disk drives and the like, which may provide non-transitory storage at any time for the software programming. All or portions of the software may at times be communicated through the Internet or various other telecommunication networks. Such communications, for example, may enable loading of the software from one computer or processor into another, for example, from a management server or host computer into the computer platform of an application server.
  • another type of media that may bear the software elements includes optical, electrical and electromagnetic waves, such as used across physical interfaces between local devices, through wired and optical landline networks and over various air-links.
  • a machine readable medium such as computer-executable code
  • a tangible storage medium such as computer-executable code
  • Non-volatile storage media include, for example, optical or magnetic disks, such as any of the storage devices in any computer(s) or the like, such as may be used to implement the databases, etc. shown in the drawings.
  • Volatile storage media include dynamic memory, such as main memory of such a computer platform.
  • Tangible transmission media include coaxial cables; copper wire and fiber optics, including the wires that comprise a bus within a computer system.
  • Carrier-wave transmission media may take the form of electric or electromagnetic signals, or acoustic or light waves such as those generated during radio frequency (RF) and infrared (IR) data communications.
  • RF radio frequency
  • IR infrared
  • Common forms of computer-readable media therefore include for example: a floppy disk, a flexible disk, hard disk, magnetic tape, any other magnetic medium, a CD-ROM, DVD or DVD-ROM, any other optical medium, punch cards paper tape, any other physical storage medium with patterns of holes, a RAM, a ROM, a PROM and EPROM, a FLASH-EPROM, any other memory chip or cartridge, a carrier wave transporting data or instructions, cables or links transporting such a carrier wave, or any other medium from which a computer may read programming code and/or data.
  • Many of these forms of computer readable media may be involved in carrying one or more sequences of one or more instructions to a processor for execution.
  • the computer system 1001 can include or be in communication with an electronic display 1035 that comprises a user interface (UI) 1040 for providing, for example, user selection of algorithms, binding measurement data, candidate peptides, and databases.
  • UI user interface
  • Examples of UIs include, without limitation, a graphical user interface (GUI) and web-based user interface.
  • Methods and systems of the present disclosure can be implemented by way of one or more algorithms.
  • An algorithm can be implemented by way of software upon execution by the central processing unit 1005 The algorithm can, for example, receive information of empirical measurements of unknown peptides in a sample, compare information of empirical
  • measurements against a database comprising a plurality of peptide sequences corresponding to candidate peptides, generate probabilities of a candidate peptide generating the observed measurement outcome set, and/or generate probabilities that candidate peptides are correctly identified in the sample.

Abstract

L'invention concerne des méthodes et systèmes pour l'identification et la quantification précises et efficaces de peptides. Selon un aspect, l'invention concerne une méthode permettant d'identifier une protéine dans un échantillon de peptides inconnus, consistant à recevoir des informations de multiples mesures empiriques réalisées sur les peptides inconnus ; comparer ces informations de mesures empiriques à une base de données comprenant de nombreuses séquences peptidiques, chaque séquence peptidique correspondant à un peptide candidat parmi une pluralité de peptides candidats ; et, pour chaque peptide candidat ou plusieurs peptides candidats parmi la pluralité de peptides candidats, produire une probabilité selon laquelle le peptide candidat génère les informations de mesures empiriques, une probabilité selon laquelle la pluralité de mesures empiriques n'est pas observée compte tenu du fait que le peptide candidat est présent dans l'échantillon ou une probabilité selon laquelle le peptide candidate est présent dans l'échantillon ; sur la base de la comparaison des informations de mesures empiriques à la base de données.
PCT/US2020/040829 2019-07-03 2020-07-03 Approches de décodage pour l'identification de protéines et de peptides WO2021003470A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US201962870283P 2019-07-03 2019-07-03
US62/870,283 2019-07-03

Publications (1)

Publication Number Publication Date
WO2021003470A1 true WO2021003470A1 (fr) 2021-01-07

Family

ID=74100775

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2020/040829 WO2021003470A1 (fr) 2019-07-03 2020-07-03 Approches de décodage pour l'identification de protéines et de peptides

Country Status (1)

Country Link
WO (1) WO2021003470A1 (fr)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11282585B2 (en) 2017-12-29 2022-03-22 Nautilus Biotechnology, Inc. Decoding approaches for protein identification
WO2023064181A1 (fr) * 2021-10-11 2023-04-20 Nautilus Biotechnology, Inc. Analyse hautement multiplexable de protéines et de protéomes
US11721412B2 (en) 2017-10-23 2023-08-08 Nautilus Subsidiary, Inc. Methods for identifying a protein in a sample of unknown proteins
US11768201B1 (en) 2016-12-01 2023-09-26 Nautilus Subsidiary, Inc. Methods of assaying proteins
WO2023212490A1 (fr) * 2022-04-25 2023-11-02 Nautilus Subsidiary, Inc. Systèmes et procédés d'évaluation et d'amélioration de la qualité de dosages moléculaires multiplex
WO2023215322A1 (fr) * 2022-05-02 2023-11-09 Merck Sharp & Dohme Llc Modélisation générative exploitant un apprentissage profond pour l'ajustement d'affinité d'anticorps

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030054408A1 (en) * 2001-04-20 2003-03-20 Ramamoorthi Ravi Methods and systems for identifying proteins
WO2019083856A1 (fr) * 2017-10-23 2019-05-02 Ignite Biosciences, Inc. Procédés et systèmes d'identification de protéines

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030054408A1 (en) * 2001-04-20 2003-03-20 Ramamoorthi Ravi Methods and systems for identifying proteins
WO2019083856A1 (fr) * 2017-10-23 2019-05-02 Ignite Biosciences, Inc. Procédés et systèmes d'identification de protéines

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
YATES, C ET AL.: "SuSPect: Enhanced Prediction of Single Amino Acid Variant (SAV) Phenotype Using Network Features", JOURNAL OF MOLECULAR BIOLOGY, 5 May 2014 (2014-05-05), pages 2692 - 2701, XP028874931, Retrieved from the Internet <URL:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4087249/?report=printable> [retrieved on 20200929], DOI: 10.1016/j.jmb. 2014.04.026 *

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11768201B1 (en) 2016-12-01 2023-09-26 Nautilus Subsidiary, Inc. Methods of assaying proteins
US11721412B2 (en) 2017-10-23 2023-08-08 Nautilus Subsidiary, Inc. Methods for identifying a protein in a sample of unknown proteins
US11282585B2 (en) 2017-12-29 2022-03-22 Nautilus Biotechnology, Inc. Decoding approaches for protein identification
US11282586B2 (en) 2017-12-29 2022-03-22 Nautilus Biotechnology, Inc. Decoding approaches for protein identification
US11545234B2 (en) 2017-12-29 2023-01-03 Nautilus Biotechnology, Inc. Decoding approaches for protein identification
WO2023064181A1 (fr) * 2021-10-11 2023-04-20 Nautilus Biotechnology, Inc. Analyse hautement multiplexable de protéines et de protéomes
WO2023212490A1 (fr) * 2022-04-25 2023-11-02 Nautilus Subsidiary, Inc. Systèmes et procédés d'évaluation et d'amélioration de la qualité de dosages moléculaires multiplex
WO2023215322A1 (fr) * 2022-05-02 2023-11-09 Merck Sharp & Dohme Llc Modélisation générative exploitant un apprentissage profond pour l'ajustement d'affinité d'anticorps

Similar Documents

Publication Publication Date Title
US11282585B2 (en) Decoding approaches for protein identification
US11721412B2 (en) Methods for identifying a protein in a sample of unknown proteins
AU2018353967B2 (en) Methods and systems for protein identification
US11754559B2 (en) Methods of assaying proteins
US11970693B2 (en) Methods of selecting binding reagents
WO2021003470A1 (fr) Approches de décodage pour l&#39;identification de protéines et de peptides
US20240158782A1 (en) Methods of selecting binding reagents

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20834511

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20834511

Country of ref document: EP

Kind code of ref document: A1