EP4377479A1 - Nachweis genetischer varianten in menschlichen leukozytenantigengenen - Google Patents
Nachweis genetischer varianten in menschlichen leukozytenantigengenenInfo
- Publication number
- EP4377479A1 EP4377479A1 EP22850393.4A EP22850393A EP4377479A1 EP 4377479 A1 EP4377479 A1 EP 4377479A1 EP 22850393 A EP22850393 A EP 22850393A EP 4377479 A1 EP4377479 A1 EP 4377479A1
- Authority
- EP
- European Patent Office
- Prior art keywords
- hla
- variants
- reads
- tumor
- allele
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 108091007433 antigens Proteins 0.000 title claims abstract description 16
- 210000000265 leukocyte Anatomy 0.000 title claims abstract description 12
- 230000002068 genetic effect Effects 0.000 title description 44
- 238000001514 detection method Methods 0.000 title description 21
- 238000000034 method Methods 0.000 claims abstract description 423
- 206010028980 Neoplasm Diseases 0.000 claims abstract description 369
- 230000008569 process Effects 0.000 claims abstract description 240
- 108090000623 proteins and genes Proteins 0.000 claims abstract description 156
- 230000000392 somatic effect Effects 0.000 claims abstract description 50
- 230000000694 effects Effects 0.000 claims abstract description 15
- 239000000427 antigen Substances 0.000 claims abstract description 14
- 102000036639 antigens Human genes 0.000 claims abstract description 14
- 108700028369 Alleles Proteins 0.000 claims description 340
- 238000012163 sequencing technique Methods 0.000 claims description 81
- 238000013507 mapping Methods 0.000 claims description 72
- 102100028972 HLA class I histocompatibility antigen, A alpha chain Human genes 0.000 claims description 43
- 108010075704 HLA-A Antigens Proteins 0.000 claims description 43
- 210000004602 germ cell Anatomy 0.000 claims description 34
- 238000001914 filtration Methods 0.000 claims description 29
- 102100028976 HLA class I histocompatibility antigen, B alpha chain Human genes 0.000 claims description 16
- 108010058607 HLA-B Antigens Proteins 0.000 claims description 16
- 238000012217 deletion Methods 0.000 claims description 14
- 230000037430 deletion Effects 0.000 claims description 14
- 230000004044 response Effects 0.000 claims description 14
- 108010052199 HLA-C Antigens Proteins 0.000 claims description 13
- 238000003780 insertion Methods 0.000 claims description 13
- 230000037431 insertion Effects 0.000 claims description 13
- 102100028971 HLA class I histocompatibility antigen, C alpha chain Human genes 0.000 claims description 10
- 230000002441 reversible effect Effects 0.000 claims description 8
- 238000007481 next generation sequencing Methods 0.000 abstract description 61
- 238000004458 analytical method Methods 0.000 abstract description 42
- 239000000523 sample Substances 0.000 description 211
- 201000011510 cancer Diseases 0.000 description 117
- 239000002773 nucleotide Substances 0.000 description 63
- 125000003729 nucleotide group Chemical group 0.000 description 62
- 210000001519 tissue Anatomy 0.000 description 56
- 210000004027 cell Anatomy 0.000 description 49
- 108020004414 DNA Proteins 0.000 description 46
- 230000014509 gene expression Effects 0.000 description 40
- 238000004422 calculation algorithm Methods 0.000 description 39
- 238000002560 therapeutic procedure Methods 0.000 description 37
- 108091032973 (ribonucleotides)n+m Proteins 0.000 description 35
- 210000002220 organoid Anatomy 0.000 description 34
- 230000035772 mutation Effects 0.000 description 33
- 101001024425 Mus musculus Ig gamma-2A chain C region secreted form Proteins 0.000 description 31
- 230000036961 partial effect Effects 0.000 description 28
- 238000012545 processing Methods 0.000 description 27
- 230000006870 function Effects 0.000 description 26
- 230000036541 health Effects 0.000 description 23
- 108020004485 Nonsense Codon Proteins 0.000 description 21
- 208000032818 Microsatellite Instability Diseases 0.000 description 19
- 108091008109 Pseudogenes Proteins 0.000 description 19
- 102000057361 Pseudogenes Human genes 0.000 description 19
- 230000037434 nonsense mutation Effects 0.000 description 19
- 230000000875 corresponding effect Effects 0.000 description 17
- 238000012549 training Methods 0.000 description 17
- 238000003556 assay Methods 0.000 description 15
- 210000004369 blood Anatomy 0.000 description 14
- 239000008280 blood Substances 0.000 description 14
- 238000009169 immunotherapy Methods 0.000 description 14
- 238000013459 approach Methods 0.000 description 13
- 102000004169 proteins and genes Human genes 0.000 description 12
- 238000012360 testing method Methods 0.000 description 12
- 238000003066 decision tree Methods 0.000 description 11
- 238000011160 research Methods 0.000 description 11
- 102000054766 genetic haplotypes Human genes 0.000 description 10
- 239000013610 patient sample Substances 0.000 description 10
- 210000003296 saliva Anatomy 0.000 description 10
- 108091028043 Nucleic acid sequence Proteins 0.000 description 9
- 230000015654 memory Effects 0.000 description 9
- 238000010200 validation analysis Methods 0.000 description 9
- 108700024394 Exon Proteins 0.000 description 8
- 108091092878 Microsatellite Proteins 0.000 description 8
- 238000010606 normalization Methods 0.000 description 8
- 208000001333 Colorectal Neoplasms Diseases 0.000 description 7
- 238000001712 DNA sequencing Methods 0.000 description 7
- 230000008901 benefit Effects 0.000 description 7
- 239000003153 chemical reaction reagent Substances 0.000 description 7
- 238000009826 distribution Methods 0.000 description 7
- 230000007246 mechanism Effects 0.000 description 7
- 238000011282 treatment Methods 0.000 description 7
- 210000004881 tumor cell Anatomy 0.000 description 7
- 108010019670 Chimeric Antigen Receptors Proteins 0.000 description 6
- 238000013528 artificial neural network Methods 0.000 description 6
- 239000012472 biological sample Substances 0.000 description 6
- 230000007717 exclusion Effects 0.000 description 6
- 238000001943 fluorescence-activated cell sorting Methods 0.000 description 6
- 238000010801 machine learning Methods 0.000 description 6
- 238000007726 management method Methods 0.000 description 6
- 230000001404 mediated effect Effects 0.000 description 6
- 230000000869 mutational effect Effects 0.000 description 6
- 210000000056 organ Anatomy 0.000 description 6
- 230000001717 pathogenic effect Effects 0.000 description 6
- 206010009944 Colon cancer Diseases 0.000 description 5
- 239000000090 biomarker Substances 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 5
- 238000004891 communication Methods 0.000 description 5
- 238000013461 design Methods 0.000 description 5
- 239000003814 drug Substances 0.000 description 5
- 230000004927 fusion Effects 0.000 description 5
- 230000002998 immunogenetic effect Effects 0.000 description 5
- 238000003908 quality control method Methods 0.000 description 5
- 102000008096 B7-H1 Antigen Human genes 0.000 description 4
- 108010074708 B7-H1 Antigen Proteins 0.000 description 4
- 102100022019 Pregnancy-specific beta-1-glycoprotein 2 Human genes 0.000 description 4
- 210000001744 T-lymphocyte Anatomy 0.000 description 4
- 230000004075 alteration Effects 0.000 description 4
- 238000013473 artificial intelligence Methods 0.000 description 4
- 230000001413 cellular effect Effects 0.000 description 4
- 230000007423 decrease Effects 0.000 description 4
- 229940079593 drug Drugs 0.000 description 4
- 230000001973 epigenetic effect Effects 0.000 description 4
- 230000017188 evasion or tolerance of host immune response Effects 0.000 description 4
- 238000007477 logistic regression Methods 0.000 description 4
- 239000000203 mixture Substances 0.000 description 4
- 244000052769 pathogen Species 0.000 description 4
- 238000002360 preparation method Methods 0.000 description 4
- 238000003860 storage Methods 0.000 description 4
- 238000007671 third-generation sequencing Methods 0.000 description 4
- GUAHPAJOXVYFON-ZETCQYMHSA-N (8S)-8-amino-7-oxononanoic acid zwitterion Chemical compound C[C@H](N)C(=O)CCCCCC(O)=O GUAHPAJOXVYFON-ZETCQYMHSA-N 0.000 description 3
- 206010069754 Acquired gene mutation Diseases 0.000 description 3
- 108700005089 MHC Class I Genes Proteins 0.000 description 3
- 238000003559 RNA-seq method Methods 0.000 description 3
- 230000003044 adaptive effect Effects 0.000 description 3
- 239000011324 bead Substances 0.000 description 3
- 238000004166 bioassay Methods 0.000 description 3
- 238000001574 biopsy Methods 0.000 description 3
- 230000004640 cellular pathway Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 210000000349 chromosome Anatomy 0.000 description 3
- 235000019506 cigar Nutrition 0.000 description 3
- 238000013145 classification model Methods 0.000 description 3
- 239000003086 colorant Substances 0.000 description 3
- 230000007812 deficiency Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 238000003745 diagnosis Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 239000012530 fluid Substances 0.000 description 3
- 239000007850 fluorescent dye Substances 0.000 description 3
- 201000005787 hematologic cancer Diseases 0.000 description 3
- 208000024200 hematopoietic and lymphoid system neoplasm Diseases 0.000 description 3
- 230000006801 homologous recombination Effects 0.000 description 3
- 238000002744 homologous recombination Methods 0.000 description 3
- 208000015181 infectious disease Diseases 0.000 description 3
- 230000008595 infiltration Effects 0.000 description 3
- 238000001764 infiltration Methods 0.000 description 3
- 229960005386 ipilimumab Drugs 0.000 description 3
- 238000011528 liquid biopsy Methods 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 210000004940 nucleus Anatomy 0.000 description 3
- 102000054765 polymorphisms of proteins Human genes 0.000 description 3
- 230000035945 sensitivity Effects 0.000 description 3
- 230000037439 somatic mutation Effects 0.000 description 3
- 230000009469 supplementation Effects 0.000 description 3
- 101150118346 HLA-A gene Proteins 0.000 description 2
- 229940076838 Immune checkpoint inhibitor Drugs 0.000 description 2
- 108700005092 MHC Class II Genes Proteins 0.000 description 2
- 108091034117 Oligonucleotide Proteins 0.000 description 2
- 238000012300 Sequence Analysis Methods 0.000 description 2
- 230000004913 activation Effects 0.000 description 2
- 230000002411 adverse Effects 0.000 description 2
- 230000003321 amplification Effects 0.000 description 2
- 230000030741 antigen processing and presentation Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 229940022399 cancer vaccine Drugs 0.000 description 2
- 238000009566 cancer vaccine Methods 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 230000002759 chromosomal effect Effects 0.000 description 2
- 238000007635 classification algorithm Methods 0.000 description 2
- 230000001186 cumulative effect Effects 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 238000000684 flow cytometry Methods 0.000 description 2
- 239000012634 fragment Substances 0.000 description 2
- 238000012252 genetic analysis Methods 0.000 description 2
- 238000009499 grossing Methods 0.000 description 2
- 239000012274 immune-checkpoint protein inhibitor Substances 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000011835 investigation Methods 0.000 description 2
- 210000004072 lung Anatomy 0.000 description 2
- 208000020816 lung neoplasm Diseases 0.000 description 2
- 230000001394 metastastic effect Effects 0.000 description 2
- 206010061289 metastatic neoplasm Diseases 0.000 description 2
- 238000002887 multiple sequence alignment Methods 0.000 description 2
- 239000013642 negative control Substances 0.000 description 2
- 229960003301 nivolumab Drugs 0.000 description 2
- 208000002154 non-small cell lung carcinoma Diseases 0.000 description 2
- 238000003199 nucleic acid amplification method Methods 0.000 description 2
- 108020004707 nucleic acids Proteins 0.000 description 2
- 102000039446 nucleic acids Human genes 0.000 description 2
- 150000007523 nucleic acids Chemical class 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000037361 pathway Effects 0.000 description 2
- 229940031734 peptide cancer vaccine Drugs 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 108090000765 processed proteins & peptides Proteins 0.000 description 2
- 238000007637 random forest analysis Methods 0.000 description 2
- 230000008707 rearrangement Effects 0.000 description 2
- 108020003175 receptors Proteins 0.000 description 2
- 102000005962 receptors Human genes 0.000 description 2
- 230000002829 reductive effect Effects 0.000 description 2
- 230000000717 retained effect Effects 0.000 description 2
- 238000013515 script Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- 230000001225 therapeutic effect Effects 0.000 description 2
- 238000013518 transcription Methods 0.000 description 2
- 230000035897 transcription Effects 0.000 description 2
- 208000029729 tumor suppressor gene on chromosome 11 Diseases 0.000 description 2
- 229960005486 vaccine Drugs 0.000 description 2
- 101150084750 1 gene Proteins 0.000 description 1
- 108091008875 B cell receptors Proteins 0.000 description 1
- 101100284398 Bos taurus BoLA-DQB gene Proteins 0.000 description 1
- 210000001266 CD8-positive T-lymphocyte Anatomy 0.000 description 1
- 229940045513 CTLA4 antagonist Drugs 0.000 description 1
- 108020004705 Codon Proteins 0.000 description 1
- 108020003215 DNA Probes Proteins 0.000 description 1
- 238000007400 DNA extraction Methods 0.000 description 1
- 239000003298 DNA probe Substances 0.000 description 1
- 206010059866 Drug resistance Diseases 0.000 description 1
- KCXVZYZYPLLWCC-UHFFFAOYSA-N EDTA Chemical compound OC(=O)CN(CC(O)=O)CCN(CC(O)=O)CC(O)=O KCXVZYZYPLLWCC-UHFFFAOYSA-N 0.000 description 1
- 108010067770 Endopeptidase K Proteins 0.000 description 1
- 241000206602 Eukaryota Species 0.000 description 1
- 108700039691 Genetic Promoter Regions Proteins 0.000 description 1
- 101000986085 Homo sapiens HLA class I histocompatibility antigen, alpha chain E Proteins 0.000 description 1
- 101000993059 Homo sapiens Hereditary hemochromatosis protein Proteins 0.000 description 1
- 101000866971 Homo sapiens Putative HLA class I histocompatibility antigen, alpha chain H Proteins 0.000 description 1
- 101001100327 Homo sapiens RNA-binding protein 45 Proteins 0.000 description 1
- 108091092195 Intron Proteins 0.000 description 1
- 238000001276 Kolmogorov–Smirnov test Methods 0.000 description 1
- 206010058467 Lung neoplasm malignant Diseases 0.000 description 1
- 101710167887 Major outer membrane protein P.IA Proteins 0.000 description 1
- 206010027476 Metastases Diseases 0.000 description 1
- 208000034179 Neoplasms, Glandular and Epithelial Diseases 0.000 description 1
- 101100384865 Neurospora crassa (strain ATCC 24698 / 74-OR23-1A / CBS 708.71 / DSM 1257 / FGSC 987) cot-1 gene Proteins 0.000 description 1
- 208000025174 PANDAS Diseases 0.000 description 1
- 208000021155 Paediatric autoimmune neuropsychiatric disorders associated with streptococcal infection Diseases 0.000 description 1
- 240000000220 Panda oleosa Species 0.000 description 1
- 235000016496 Panda oleosa Nutrition 0.000 description 1
- ZYFVNVRFVHJEIU-UHFFFAOYSA-N PicoGreen Chemical compound CN(C)CCCN(CCCN(C)C)C1=CC(=CC2=[N+](C3=CC=CC=C3S2)C)C2=CC=CC=C2N1C1=CC=CC=C1 ZYFVNVRFVHJEIU-UHFFFAOYSA-N 0.000 description 1
- 102100031498 Putative HLA class I histocompatibility antigen, alpha chain H Human genes 0.000 description 1
- 102100038823 RNA-binding protein 45 Human genes 0.000 description 1
- 206010041067 Small cell lung cancer Diseases 0.000 description 1
- 238000000692 Student's t-test Methods 0.000 description 1
- 230000006044 T cell activation Effects 0.000 description 1
- 108091008874 T cell receptors Proteins 0.000 description 1
- 102000016266 T-Cell Antigen Receptors Human genes 0.000 description 1
- 101150049278 US20 gene Proteins 0.000 description 1
- 101150114976 US21 gene Proteins 0.000 description 1
- 108020005202 Viral DNA Proteins 0.000 description 1
- 101500012027 Viscum album Beta-galactoside-specific lectin 1 chain A isoform 1 Proteins 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000003213 activating effect Effects 0.000 description 1
- 239000012190 activator Substances 0.000 description 1
- 238000007792 addition Methods 0.000 description 1
- 230000004931 aggregating effect Effects 0.000 description 1
- 125000003275 alpha amino acid group Chemical group 0.000 description 1
- 150000001413 amino acids Chemical class 0.000 description 1
- 238000012197 amplification kit Methods 0.000 description 1
- 108010038083 amyloid fibril protein AS-SAM Proteins 0.000 description 1
- 229960003852 atezolizumab Drugs 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000007321 biological mechanism Effects 0.000 description 1
- 230000000903 blocking effect Effects 0.000 description 1
- 210000000601 blood cell Anatomy 0.000 description 1
- 210000001124 body fluid Anatomy 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000022534 cell killing Effects 0.000 description 1
- 238000002659 cell therapy Methods 0.000 description 1
- 108091092240 circulating cell-free DNA Proteins 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000013264 cohort analysis Methods 0.000 description 1
- 238000002648 combination therapy Methods 0.000 description 1
- 239000002299 complementary DNA Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000011109 contamination Methods 0.000 description 1
- 239000013068 control sample Substances 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 210000001151 cytotoxic T lymphocyte Anatomy 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000002405 diagnostic procedure Methods 0.000 description 1
- 238000007877 drug screening Methods 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 230000008482 dysregulation Effects 0.000 description 1
- -1 e.g. Proteins 0.000 description 1
- 230000002357 endometrial effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000001747 exhibiting effect Effects 0.000 description 1
- 238000013401 experimental design Methods 0.000 description 1
- 238000007672 fourth generation sequencing Methods 0.000 description 1
- 231100000221 frame shift mutation induction Toxicity 0.000 description 1
- 230000037433 frameshift Effects 0.000 description 1
- 108020001507 fusion proteins Proteins 0.000 description 1
- 102000037865 fusion proteins Human genes 0.000 description 1
- 238000001502 gel electrophoresis Methods 0.000 description 1
- 238000012268 genome sequencing Methods 0.000 description 1
- 230000012010 growth Effects 0.000 description 1
- 238000009396 hybridization Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000005965 immune activity Effects 0.000 description 1
- 210000002865 immune cell Anatomy 0.000 description 1
- 230000005746 immune checkpoint blockade Effects 0.000 description 1
- 230000028993 immune response Effects 0.000 description 1
- 230000037451 immune surveillance Effects 0.000 description 1
- 210000000987 immune system Anatomy 0.000 description 1
- 238000000338 in vitro Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000002147 killing effect Effects 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 238000009092 lines of therapy Methods 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 239000006193 liquid solution Substances 0.000 description 1
- 230000004777 loss-of-function mutation Effects 0.000 description 1
- 201000005202 lung cancer Diseases 0.000 description 1
- 208000037841 lung tumor Diseases 0.000 description 1
- 230000013011 mating Effects 0.000 description 1
- 238000002483 medication Methods 0.000 description 1
- 108020004999 messenger RNA Proteins 0.000 description 1
- 201000001997 microphthalmia with limb anomalies Diseases 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 108091027963 non-coding RNA Proteins 0.000 description 1
- 102000042567 non-coding RNA Human genes 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 239000012188 paraffin wax Substances 0.000 description 1
- 229960002621 pembrolizumab Drugs 0.000 description 1
- 210000003819 peripheral blood mononuclear cell Anatomy 0.000 description 1
- 229950010773 pidilizumab Drugs 0.000 description 1
- 230000003389 potentiating effect Effects 0.000 description 1
- 238000000513 principal component analysis Methods 0.000 description 1
- 102000004196 processed proteins & peptides Human genes 0.000 description 1
- 210000002307 prostate Anatomy 0.000 description 1
- 238000011002 quantification Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000007670 refining Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000012552 review Methods 0.000 description 1
- 238000007480 sanger sequencing Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000002864 sequence alignment Methods 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 230000008054 signal transmission Effects 0.000 description 1
- 208000000587 small cell lung carcinoma Diseases 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000010186 staining Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 230000004083 survival effect Effects 0.000 description 1
- 238000003239 susceptibility assay Methods 0.000 description 1
- 238000002626 targeted therapy Methods 0.000 description 1
- 230000005945 translocation Effects 0.000 description 1
- 229950007217 tremelimumab Drugs 0.000 description 1
- 238000009966 trimming Methods 0.000 description 1
- 238000011144 upstream manufacturing Methods 0.000 description 1
- 210000002700 urine Anatomy 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
- 238000007482 whole exome sequencing Methods 0.000 description 1
- 238000012070 whole genome sequencing analysis Methods 0.000 description 1
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
- G16B20/00—ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations
- G16B20/20—Allele or variant detection, e.g. single nucleotide polymorphism [SNP] detection
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H15/00—ICT specially adapted for medical reports, e.g. generation or transmission thereof
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
-
- C—CHEMISTRY; METALLURGY
- C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
- C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
- C12Q1/00—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions
- C12Q1/68—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions involving nucleic acids
- C12Q1/6876—Nucleic acid products used in the analysis of nucleic acids, e.g. primers or probes
- C12Q1/6881—Nucleic acid products used in the analysis of nucleic acids, e.g. primers or probes for tissue or cell typing, e.g. human leukocyte antigen [HLA] probes
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/20—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for electronic clinical trials or questionnaires
Definitions
- HLA Human Leukocyte Antigen Class I
- a computer-implemented method of identifying somatic variants in a human leukocyte antigen (HLA) gene in a subject includes: for an HLA gene, obtaining a patient specific HLA reference genome from the HLA type data; receiving HLA sequence read data from one or more normal samples and receiving HLA sequence read data for a tumor sample associated with a patient; aligning the HLA sequence read data for the one or more normal samples and the HLA sequence data for the tumor sample to the patient specific HLA reference genome; adjusting a mapping quality score associated with each read having a low mapping quality score indicative of multiple-mapping; performing a variant calling process on the filtered aligned read data and determining the somatic variants of the HLA class; and generating and storing a report of somatic variants of the HLA gene annotated for functional effect.
- the method further includes: before determining the somatic variants of the HLA class, removing variants resulting from the variant calling process that are
- the method further includes removing the variants by: identifying, from a list of candidate somatic variants, variants in which all sequencing reads that support the variant map to the right of the variant or to the left of the variant.
- the method further includes if the all the sequence reads map to the right or to the left of the variant, determining if the sequencing reads map to a different HLA gene and filtering the sequencing reads that map to the different HLA gene.
- the method further includes before determining the somatic variants of the HLA class, removing germline variants from variants identified by the variant calling process.
- the method further includes removing the germline variants by: removing the germline variants from a list of candidate somatic variants, in response to tumor sample reads being below a threshold value.
- the method further includes removing the germline variants by: removing the germline variants from a list of candidate somatic variants, in response to normal sample reads being above a threshold value.
- the method further includes removing the germline variants by: removing the germline variants from a list of candidate somatic variants, in response to tumor variant allele frequency (VAF) and normal VAF being equal.
- VAF tumor variant allele frequency
- the method further includes removing the germline variants by: removing the germline variants from a list of candidate somatic variants, in response to a ratio of tumor variant allele frequency (VAF) to normal VAF being lower than a threshold value.
- VAF tumor variant allele frequency
- the method further includes before determining the somatic variants of the HLA class, removing variants likely resulting from sequencing artifacts from variants identified by the variant calling process.
- the method further includes removing variants likely resulting from sequencing artifacts from variants identified by the variant calling process by: removing variants from a list of candidate somatic variants, if all sequencing reads that support the variant map only to the forward strand or only to the reverse strand.
- the method further includes designating a variant as a somatic variant, if the ratio of tumor variant allele frequency (VAF) to normal VAF is greater than a threshold value.
- VAF tumor variant allele frequency
- the FILA class is FILA Class 1
- the somatic variants are located in one of the following genes: FILA-A, FILA-B, and FILA-C.
- the method further includes obtaining FILA type data that includes FILA reference sequence data corresponding to a plurality of different FILA alleles.
- the method further includes updating the patient specific FILA reference genome to match the FILA sequence read data of the one or more normal samples.
- the one or more normal samples are associated with the patient.
- the one or more normal samples are selected from normal samples of non-patients determined to have a matching an FILA genotype of the patient.
- the method further comprises obtaining the patient specific FILA reference genome by applying the FILA type data to a FILA typing module and generating, using the FILA typing module, the patient specific FILA reference genome.
- the method further comprises, prior to performing the variant calling process on the filtered aligned read data, removing non-paired reads from the filtered aligned read data.
- removing non-paired reads from the filtered aligned read data comprises removing read data having an unmapped read end.
- adjusting the mapping quality score associated with each remaining read having the low mapping quality score indicative of multiple-mapping comprises setting an equal weight for each remaining read having the low mapping quality score indicative of multiple mapping.
- a computer-implemented method of detecting loss of heterozygosity (LOH) of a human leukocyte antigen (FILA) gene in a patient includes: obtaining FILA coverage feature metrics of a biological sample; providing one or more of the HLA coverage feature metrics to a three-class HLA loss of heterozygosity (LOH) modeling process trained to classify the biological sample as corresponding to one of three LOH classes, no LOH, partial LOH, or clonal LOH and determining the LOH class of the sample and determining, using the three-class HLA LOH modeling process, the LOH class for the HLA gene; and generating and storing a report of the determined LOH class for the HLA gene, wherein the obtaining HLA coverage feature metrics of the biological sample comprises, receiving normal next generation sequencing data generated from a panel of normal samples of non-patients; aligning the next generation sequencing data against a reference genome to determine a normal mapped reads dataset and a normal unmapped
- FIG. 1 illustrates an example workflow 10 for next generation sequencing, bioinformatics processing, and report generation, in an example.
- FIG. 2 illustrates a schematic of an example process for Human Leukocyte Antigen Class I (HLA) detection and analysis.
- FIG. 3 illustrates an example process schematic for data flow for an HLA typing model and a loss of heterozygosity (LOH) in HLA genes (LOH) model (collectively the HLA and HLA- LOH model).
- HLA Human Leukocyte Antigen Class I
- FIG. 4 illustrates an example HLA typing report, generated in an example.
- FIGS. 5A, 5B, and 5C collectively illustrate plots of coverage metrics calculated for different examples of the techniques herein, some in comparison to non-technique examples, and some without the filter steps.
- FIG. 5A shows data that were calculated using all disclosed steps and features
- FIG. 5B shows data calculated without aligning discarded/unmapped reads to HLA genes
- FIG. 5C shows data calculated without replacing the HLA reference sequences with the variants detected in the sequence data generated by the patient sample.
- Light colors (lighter blue and lighter red) indicate areas of low coverage and black dots indicate positions where the sequences of the two alleles diverge from one another.
- FIG. 6 illustrates an example shallow decision tree showing the use of coverage metrics to predict HLA- LOH.
- FIGS. 7A and 7B collectively illustrate the results of an optional biological assay used to validate the predictions of the HLA and LOH model.
- FIGS. 8A, 8B, and 8C collectively illustrate coverage metrics plots calculated by the methods disclosed herein for different types of tissues.
- FIG. 8A shows coverage data calculated for the non-cancer sample.
- FIG. 8B shows coverage data calculated for the cancer sample tissue extracted from the same patient as the non-cancer sample.
- FIG. 8A shows coverage data calculated for the non-cancer sample.
- 8C shows coverage data for a tumor organoid derived from the cancer sample tissue.
- FIGS. 9A, 9B, 9C, and 9D collectively illustrate how various model features lead to more robust alignments and less noisy signal for downstream analysis by comparing plots of coverage metrics calculated for different examples of the techniques herein with coverage metrics calculated for non-technique examples, and some without the filter steps.
- FIG. 10 illustrates an example system for HLA and HLA-LOH analysis that may be implemented on a network accessible processing system for performing the processes described herein.
- FIG. 11 illustrates how HLA-LOH can potentially lead to escape of immune pressure.
- FIG. 12 illustrates relative differences in allele coverage metrics calculated in order to detect HLA-LOH, including B allele frequencies (BAF) and Log Coverage ratios, between the Tumor and Normal sample.
- BAF B allele frequencies
- Log Coverage ratios between the Tumor and Normal sample.
- the cancer specimen analyzed for these results represents a strong HLA-LOH.
- the allele predicted to have been lost and the allele predicted to be stable are highlighted in red and blue, respectively.
- Light colors (light blue and light red) indicate areas of low coverage and black dots indicate positions where the sequences of the two alleles diverge from one another.
- FIG. 13 is a table showing the percent and number of samples in the xT 500 cohort predicted to have HLA-LOH by the model, categorized by cancer type.
- FIG. 14 illustrates predicted HLA-LOH status among all samples in the xT 500 cohort.
- Each column represents a sample, with the LOH status of each HLA gene (HLA-A, HLA-B, or HLA-C as denoted by the y-axis label) shown as Predicted LOH (red), Predicted Stable (blue), or Homozygous (grey).
- FIG. 15 illustrates the association or lack of association between Tumor Mutational Burden (TMB) and LOH status. These charts compare the log normalized TMB between samples with no HLA-LOH (blue) and predicted HLA-LOH (red). Significance was determined by Student’s T test.
- FIG. 16 is a schematic of an example process for determining HLA LOH status in a three-class classification process having two classification stages.
- FIG. 17 is a schematic of an example HLA LOH classification stage of FIG. 16.
- FIG. 18 illustrates normal sample plots of (i) read coverage (number of reads) on the y- axis for two different alleles (B * 44:02 (red data points) and B * 07:02 (blue data points)) as a function of nucleotide position, (ii) BAF for the two different alleles as a function of nucleotide position, and (iii) Log Ratio of read coverage in the tumor sample to the read coverage in the normal sample as a function of nucleotide position.
- FIG. 19 illustrates tumor sample plots of (i) read coverage (number of reads) on the y- axis for two different alleles (B * 44:02 (red data points) and B * 07:02 (blue data points)) as a function of nucleotide position, (ii) BAF for the two different alleles as a function of nucleotide position, and (iii) Difference between Log Ratio of the two different alleles as a function of nucleotide position, illustrating a partial LOH example.
- FIG. 20 illustrates normal and tumor sample plots of (i) read coverage (number of reads) on the y-axis for two different alleles (B * 44:02 (red data points) and B * 07:02 (blue data points)) as a function of nucleotide position, (ii) Log Ratio of read coverage in the tumor sample to the read coverage in the normal sample as a function of nucleotide position, and (iii) Difference between Log Ratio of the two different alleles as a function of nucleotide position, illustrating a clonal LOH example.
- FIG. 21 illustrates a schematic of an example process for Human Leukocyte Antigen (HLA) Class I genetic variant detection and analysis.
- HLA Human Leukocyte Antigen
- FIG. 22 illustrates a plot of the Spearman correlation coefficient between the variant allele frequency (VAF) associated with the variant detected with an example implementation of the process of FIG. 21 (x-axis) and the VAF associated with the variant detected by a conventional technique, Polysolver (y-axis).
- VAF variant allele frequency
- FIG. 23 illustrates a bar graph comparing expression levels (y-axis) between an allele with PTC and an allele without PTC for the same HLA gene in six different TCGA samples (x- axis).
- FIG. 24 illustrates a bar graph comparing expression levels (y-axis) between different alleles (x-axis) for an example sample, TCGA-CV-6937.
- FIG. 25 illustrates HLA expression data, in particular RNA allelic imbalance, for different TCGA sample data, obtained using techniques herein and indicating gene expression profiles for nonsense mutations (NM), no identified nonsense mutations (No NM), loss of heterozygosity (LOH), and no LOH.
- NM nonsense mutations
- No NM no identified nonsense mutations
- LOH loss of heterozygosity
- FIG. 26 illustrates plots of different metrics for a panel of normal samples (left) and a patient tumor sample (right) for a stable A * 02:01/A * 03:01 sample.
- Wide lines represent the coverage, B-allele fraction or log ratio coverage in the normal (left) or tumor (right) sample.
- the thin lines represent the coverage of the panel of normals and B-allele fraction or log ratio coverage calculated using the panel of normals
- FIG. 27 illustrates a plot of B-allele fraction of the normal and tumor sample in -6,000 HLA-A * 02:01 heterozygous samples, showing allele LOH, stable, clonal loss, and partial loss.
- FIG. 28 illustrates a distribution plot of B-allele fraction in normal samples.
- FIG. 29 illustrates a frequency plot of partial loss calls by B-allele fraction percentile (each bin contains 600 samples).
- FIG. 30 illustrates a frequency plot of clonal loss calls by B-allele fraction percentile (each bin contains 600 samples).
- FIG. 31 is a plot showing the median B-allele fraction of HLA-A * 02:01 in the normal sample when paired with other frequent alleles (> 30 samples).
- FIGS. 32A and 32B illustrate the BAF of HLA-A * 02:01 by position in the HLA-A * 02:01 / A:03:01 haplotype, with FIG. 32A for normal samples and FIG. 32B for tumor samples.
- FIG. 33 illustrates a median BAF and 95% confidence interval of BAF of HLA-A * 02:01 by position in the HLA-A * 02:01 / A:03:01 haplotype
- FIGS. 34A and 34B illustrate FILA-A LOFI status in function of the BAF of A * 02:01 in the normal and tumor sample in the 6,005 A * 02:01 heterozygous samples when using the patient’s normal (FIG. 34A) or the panel of normals (FIG. 34B).
- FIGS. 35A-35D illustrate a frequency plots of partial loss (FIGS. 34a and 34B) or clonal loss (FIGS. 34C and 34D) calls by B-allele fraction percentile (each bin contains 600 samples).
- “Pseudogene” means a non-functional HLA gene (for example, HLA-Y) and/or an HLA gene that isn’t expressed. HLA pseudogenes may not impact a patient’s health, immune system activity and/or control of cancer cells, but these pseudogenes may have genetic sequences that are similar to the genetic sequences of functional HLA genes, such that sequence reads from HLA pseudogenes could potentially align to functional HLA genes.
- Genetic analyzer means a device, system, and/or methods for determining the characteristics (including sequences) of nucleic acid molecules (including DNA, RNA, etc.) present in biological specimens (including tumors, biopsies, tumor organoids, blood samples, saliva samples, or other tissues or fluids).
- “Targeted Panel” means a combination of probes for next-generation sequencing of a patient’s biological specimens (including tumors, biopsies, tumor organoids, blood samples, saliva samples, or other tissues or fluids) which are selected to map one or more loci on one or more chromosomes.
- “Sequencing probe” means a collection of chemicals which attach to a locus of a chromosome based on the expected sequence of nucleotides at the RNA or DNA present at that locus.
- RNA read count means the read counts of RNA or cDNA generated from a genetic analyzer.
- Bioinformatics pipeline means a series of processing stages of a pipeline to instantiate bioinformatics reporting regarding next-generation sequencing results of a patient’s tumor or normal tissue or bodily fluids to extract and report on variants present in the patient’s genome.
- Genetic profile means a combination of one or more variants, RNA transcriptomes, or other informative genetic characteristics determined for a patient from next-generation sequencing.
- Genetic sequence means a recordation of a series of nucleotides present in a patient’s RNA or DNA as determined from sequencing the patient’s tissue or fluids.
- Variant means a difference in a genetic sequence or genetic profile when compared to a reference genetic sequence or expected genetic profile.
- “Expression level” means the number of copies of an RNA or protein molecule generated by a gene or other genetic locus, which may be defined by a chromosomal location or other genetic mapping indicator.
- Gene product means a molecule (including a protein or RNA molecule) generated by the manipulation (including transcription) of the gene or other genetic locus, which may be defined by a chromosomal location or other genetic mapping indicator.
- NGS DNA Next-Generation Sequencing
- Class I genes include HLA-A, -B, and -C, as well as the non- classical MHC-lb genes HLA-E, -F, and -G.
- Class II genes include DRA, DRB1 , DQA1 , DQB1 , DPA1 , and DPB1. Multiple alleles exist for each genetic locus.
- HLA polymorphic lipoprotein
- the polymorphic nature of HLA is an important evolutionary development, as it allows the population to display a wide range of antigens to the immune system.
- the large degree of polymorphism at the Class I and Class II loci poses a significant challenge for detecting mutation and loss of heterozygosity.
- the instant disclosure provides methods and systems for overcoming the limitations associated with NGS to efficiently and accurately detect mutation and/or loss of heterozygosity (LOH) of HLA genes (also termed “HLA-LOH” herein) in a subject, especially in cancer cells within a subject.
- LHO heterozygosity
- HLA-LOH and/or HLA mutations may occur in cancer cells without occurring in the healthy/non-cancer cells in a subject.
- the HLA-LOH and/or HLA variant calling processes herein may be executed on one or more network accessible computer processing systems, including network accessible devices communicatively coupled to other computer systems, such as other NGS systems.
- the processes include, initially receiving genetic material (DNA or RNA) isolated from a patient specimen and sequenced, for example, using a NGS technique.
- the processes may receive only the sequence data.
- the specimen may be any biological sample obtained from the patient, such as a tissue sample (e.g., tumor tissue from a biopsy), a cell sample, blood, saliva, urine, and the like.
- Both cancer and non-cancer specimens may be isolated and sequenced by the computer processing systems performing the patient-specific HLA gene sequence analysis, and such systems may store the sequence data in a set of data files for the cancer specimens and a set of data files for non-cancer specimens. Each file may be configured to store the sequence of each detected read and the number of times (counts) that a sequence was detected.
- Example data file formats include a BCL file or a FASTQ file, where the FASTQ format further includes a quality score for each read.
- the computer processing systems may pre-process the sequence data by filtering and/or cleaning the data and align that pre-processed data against a reference genome, for example, using a bioinformatics pipeline executed using the computer processing system.
- the reference genome build is the hg 19 genome (see, e.g., GenBank assembly accession: GCA 000001405.1).
- GenBank assembly accession: GCA 000001405.1 the genetic sequence of HLA genes there can be considerable variety from person to person, however the hg 19 genome contains only one allele for each HLA gene; therefore many patient sequencing reads detected from the HLA genes may not map to hg 19.
- the normalization and alignment for sequence data occurs for both cancer and non-cancer specimens, yielding a set of output files for cancer specimens and a set of output files for non-cancer specimens.
- the output files may store genetic positions indicating the location in the reference genome that matches the sequence of each read, and additional information relating to mapping attributes and mapping quality of each read.
- Example file formats include a Binary Alignment Map (BAM) file.
- BAM Binary Alignment Map
- the process generates normal tissue BAM files and tumor tissue BAM files.
- Unmapped reads that is, reads that do not match the genome with quality scores that exceed quality thresholds, are stored in the BAM file with corresponding read flags indicating that the read did not map successfully. This may be due to high numbers of mismatched bases or a high degree of multimapping. In some examples, reads bearing this unmapped flag are generally excluded from downstream analysis (variant calling for non HLA genes, etc.).
- FIG. 1 illustrates an example workflow 10 for next generation sequencing, bioinformatics processing, and report generation, in an example.
- cancer samples and non-cancer samples may be processed by DNA next generation sequencing (NGS) 12, designed to sequence either the whole exome or a targeted panel of cancer-related genes, to generate DNA sequencing data, and the DNA sequencing data may be processed by a bioinformatics pipeline 14 to generate FILA-LOFI results (among other outputs) for each sample.
- the cancer sample may be a tissue sample or blood sample containing cancer cells.
- a tumor organoid sample may be processed instead of the patient cancer sample.
- germline (“normal”, non-cancerous) DNA may be extracted from either blood (for example, if a patient has cancer that is not a blood cancer) or saliva (for example, if a patient has blood cancer).
- Normal blood samples may be collected from patients (for example, in PAXgene Blood DNA Tubes) and saliva samples may be collected from patients (for example, in Oragene DNA Saliva Kits).
- Blood cancer samples may be collected from patients (for example, in EDTA collection tubes).
- Macrodissected FFPE tissue sections (which may be mounted on a histopathology slide) from solid tumor samples may be analyzed by pathologists to determine overall tumor amount in the sample and percent tumor cellularity as a ratio of tumor to normal nuclei.
- background tissue may be excluded or removed such that the section meets a tumor purity threshold (in one example, at least 20% of the nuclei in the section are tumor nuclei).
- DNA may be isolated from blood samples, saliva samples, and tissue sections using commercially available reagents, including proteinase K to generate a liquid solution of DNA.
- reagents including proteinase K to generate a liquid solution of DNA.
- Each solution of isolated DNA may be subjected to a quality control protocol to determine the concentration and/or quantity of the DNA molecules in the solution, which may include the use of a fluorescent dye and a fluorescence microplate reader, standard spectrofluorometer, or filter fluorometer.
- isolated DNA molecules may be mechanically sheared to an average length using an ultrasonicator (for example, a Covaris ultrasonicator).
- the DNA molecules may also be analyzed to determine their fragment size, which may be done through gel electrophoresis techniques and may include the use of a device such as a LabChip GX Touch.
- DNA libraries may be prepared from the isolated DNA, for example, using the KAPA Hyper Prep Kit, a New England Biolabs (NEB) kit, or a similar kit.
- DNA library preparation may include the ligation of adapters onto the DNA molecules.
- UDI adapters including Roche SeqCap dual end adapters, or UMI adapters (for example, full length or stubby Y adapters) may be ligated to the DNA molecules.
- adapters are nucleic acid molecules that may serve as barcodes to identify DNA molecules according to the sample from which they were derived and/or to facilitate the downstream bioinformatics processing and/or the next generation sequencing reaction.
- the sequence of nucleotides in the adapters may be specific to a sample in order to distinguish samples.
- the adapters may facilitate the binding of the DNA molecules to anchor oligonucleotide molecules on the sequencer flow cell and may serve as a seed for the sequencing process by providing a starting point for the sequencing reaction.
- DNA libraries may be amplified and purified using reagents, for example, Axygen MAG PCR clean up beads. Then the concentration and/or quantity of the DNA molecules may be quantified using a fluorescent dye and a fluorescence microplate reader, standard spectrofluorometer, or filter fluorometer.
- DNA libraries may be pooled (two or more DNA libraries may be mixed to create a pool) and treated with reagents to reduce off-target capture, for example Human COT-1 and/or IDT xGen Universal Blockers. Pools may be dried in a vacufuge and resuspended. DNA libraries or pools may be hybridized to a probe set (for example, a probe set specific to a panel that includes approximately 100, 600, 1,000, 10,000, etc.
- a probe set for example, a probe set specific to a panel that includes approximately 100, 600, 1,000, 10,000, etc.
- IDT xGen Exome Research Panel v1.0 probes IDT xGen Exome Research Panel v2.0 probes, other IDT probe panels, Roche probe panels, another probe panel that captures the human exome, or another probe panel
- amplified with commercially available reagents for example, the KAPA HiFi HotStart ReadyMix.
- Pools may be incubated in an incubator, PCR machine, water bath, or other temperature modulating device to allow probes to hybridize. Pools may then be mixed with Streptavidin- coated beads or another means for capturing hybridized DNA-probe molecules, especially DNA molecules representing exons of the human genome and/or genes selected for a genetic panel.
- Pools may be amplified and purified more than once using commercially available reagents, for example, the KAPA HiFi Library Amplification kit and Axygen MAG PCR clean up beads, respectively.
- the pools or DNA libraries may be analyzed to determine the concentration or quantity of DNA molecules, for example by using a fluorescent dye (for example, PicoGreen pool quantification) and a fluorescence microplate reader, standard spectrofluorometer, or filter fluorometer.
- the DNA library preparation and/or whole exome capture steps of the process 12 may be performed partially or wholly with an automated system, using a liquid handling robot (for example, a SciClone NGSx).
- a liquid handling robot for example, a SciClone NGSx.
- the library amplification may be performed on a device, for example, an lllumina C-Bot2, and the resulting flow cell containing amplified target-captured DNA libraries may be sequenced on a next generation sequencer, for example, an lllumina HiSeq 4000 or an lllumina NovaSeq 6000 to a unique on-target depth selected by the user, for example, 300x, 400x, 500x, 10,000x, etc. Samples may be further assessed for uniformity with each sample required to have 95% of all targeted bp sequenced to a minimum depth selected by the user, for example, 300x.
- the next generation sequencer may generate a FASTQ, BCL, or other file for each flow cell or each patient sample.
- the bioinformatics pipeline 14 may filter FASTQ data obtained from the NGS Lab process 12.
- Filtering FASTQ data may include correcting sequencer errors and removing (trimming) low quality sequences or bases, adapter sequences, contaminations, chimeric reads, overrepresented sequences, biases caused by library preparation, amplification, or capture, and other errors.
- Entire reads, individual nucleotides, or multiple nucleotides that are likely to have errors may be discarded based on the quality rating associated with the read in the FASTQ file, the known error rate of the sequencer, and/or a comparison between each nucleotide in the read and one or more nucleotides in other reads that has been aligned to the same location in the reference genome. Filtering may be done in part or in its entirety by
- FASTQ files may be analyzed for rapid assessment of quality control and reads, for example, by a sequencing data QC software such as AfterQC, Kraken, RNA-SeQC, FastQC, (see lllumina, BaseSpace Labs or illumina.com/products/by-type/informatics-products/basespace-sequence- hub/apps/fastqc.html), or another similar software program.
- a sequencing data QC software such as AfterQC, Kraken, RNA-SeQC, FastQC, (see lllumina, BaseSpace Labs or illumina.com/products/by-type/informatics-products/basespace-sequence- hub/apps/fastqc.html), or another similar software program.
- paired-end reads reads may be merged.
- each read in the file may be aligned to the location in the human genome having a sequence that best matches the sequence of nucleotides in the read.
- There are many software programs designed to align reads for example, Novoalign (Novocraft, Inc.), Bowtie, Burrows Wheeler Aligner (BWA), programs that use a Smith-Waterman algorithm, etc.
- Alignment may be directed using a reference genome (for example, hg 19, GRCh38, hg38, GRCh37, other reference genomes developed by the Genome Reference Consortium, etc.) by comparing the nucleotide sequences in each read with portions of the nucleotide sequence in the reference genome to determine the portion of the reference genome sequence that is most likely to correspond to the sequence in the read.
- the alignment may generate a SAM file, which stores the locations of the start and end of each read according to coordinates in the reference genome and the coverage (number of reads) for each nucleotide in the reference genome.
- the SAM files may be converted to BAM files, BAM files may be sorted, and duplicate reads may be marked for deletion, resulting in de-duplicated BAM files.
- a BAM file may contain reads from both a cancer sample and a normal sample, and these samples may be derived from the same patient.
- a matched tumor-normal oncology targeted panel single-site Next Generation Sequencing (NGS) assay may be used for pre-processing.
- the assay is a laboratory-developed test (LDT).
- the assay is a marketed assay approved by a regulatory body.
- the assay may include reagents, software, instruments, and procedures for testing DNA extracted from formalin-fixed, paraffin-embedded (FFPE) tumor specimens and matched normal blood or saliva specimens.
- FFPE formalin-fixed, paraffin-embedded
- the assay is designed to detect and identify somatic alterations for use and interpretation by qualified healthcare professionals to aid in the clinical management of previously diagnosed cancer patients with solid malignant neoplasms.
- the assay is a next generation sequencing-based in vitro diagnostic device intended for use in the detection of substitutions (single nucleotide variants (SNVs) and multi-nucleotide variants (MNVs)) and insertion and deletion alterations (INDELs) in 648 genes, as well as microsatellite instability (MSI) status using DNA isolated from formalin- fixed paraffin embedded (FFPE) tumor tissue specimens, and matched normal specimens, from previously diagnosed cancer patients.
- the assay may provide tumor mutation profiling to be used by qualified health care professionals in accordance with professional guidelines in oncology for patients with malignant neoplasms.
- the assay workflow includes sample processing through to the completion of sequencing and creation of an aligned BAM file for patient-matched tumor and normal samples.
- HLA-LOH determination involves novel bioinformatics pipeline software to add a parallel analysis of sequencing results to support HLA-LOH determination.
- the sequencing assay includes DNA extraction from FFPE tissue samples and matched normal saliva or blood samples. Extracted DNA undergoes whole- genome shotgun library construction and hybridization-based capture of specified regions from 648 cancer-related genes (including intronic overhang and selected promoter regions), 196 loci for microsatellite instability (MSI), and the sequencing probes also include probes specifically designed to efficiently capture a diverse array of HLA alleles.
- MSI microsatellite instability
- HLA-LOH and/or HLA variants may be used to determine whether a patient sample has HLA-LOH and/or HLA variants, for example.
- BAM files may be analyzed to detect genetic variants, including single nucleotide variants (SNVs), copy number variants (CNVs), gene rearrangements, etc.
- SNVs single nucleotide variants
- CNVs copy number variants
- VCF variant call format
- de-duplicated BAM files and a VCF generated from the variant calling pipeline may be used to compute read depth and variation in heterozygous germline SNVs between the tumor and normal samples (or between the tumor sample and a pool of process matched normal controls for tumor- only cases when the matched normal sample is not available).
- Circular binary segmentation may be applied and segments may be selected with highly differential log2 ratios between the tumor and its comparator (matched normal or normal pool).
- Approximate integer copy number may be assessed from a combination of differential coverage in segmented regions and an estimate of stromal admixture (for example, tumor purity, or the portion of a sample that is tumor vs. non-tumor) generated by analysis of heterozygous germline SNVs.
- the copy number status of chromosome (chr) 6 and/or arms or other portions of chr 6 in the tumor sample and/or the normal sample may be detected by the bioinformatics pipeline and/or received by the systems and methods.
- tumor FASTQ files may be aligned against the human reference genome using BWA for DNA files.
- DNA reads may be sorted and duplicates may be marked with a software, for example, SAM Blaster. Discordant and split reads may be further identified and separated. These data may be read into a software, for example, LUMPY, for structural variant detection.
- Structural alterations may be grouped by type, recurrence, and presence and stored within a database and displayed through a fusion viewer software tool.
- the fusion viewer software tool may reference a database, for example, Ensembl, to determine the gene and proximal exons surrounding the breakpoint for any possible transcript generated across the breakpoint.
- the fusion viewer tool may then place the breakpoint 5’ or 3’ to the subsequent exon in the direction of transcription. For inversions, this orientation may be reversed for the inverted gene.
- the translated amino acid sequences may be generated for both genes in the chimeric protein, and a plot may be generated containing the remaining functional domains for each protein, as returned from a database, for example, Uniprot.
- a report generation process 16 may be used for variant classification and reporting.
- the process 16 may detect variants and investigate detected variants following criteria from known evolutionary models, functional data, clinical data, literature, and other research endeavors, including tumor organoid experiments.
- variants may be prioritized and classified based on known gene-disease relationships, hotspot regions within genes, internal and external somatic databases, primary literature, and other features of somatic drivers.
- Variants may be added to a patient (or sample, for example, organoid sample) report based on recommendations from the AMP/ASCO/CAP guidelines. Additional guidelines may be followed. Briefly, pathogenic variants with therapeutic, diagnostic, or prognostic significance may be prioritized in the report.
- Non-actionable pathogenic variants may be included as biologically relevant, followed by variants of uncertain significance. Translocations may be reported based on features of known gene fusions, relevant breakpoints, and biological relevance. Evidence may be curated from public and private databases or research and presented as 1) consensus guidelines 2) clinical research, or 3) case studies, with a link to the supporting literature. Germline alterations may be reported as secondary findings in a subset of genes for consenting patients. These may include genes recommended by the ACMG and additional genes associated with cancer predisposition or drug resistance.
- microsatellite instability status For detecting microsatellite instability status (MSI), the probes used during library preparation before sequencing may target microsatellite regions (for example, approximately 40, 50, 60, 100, 1,000 regions).
- a MSI classification algorithm classifies tumors into three categories: microsatellite instability-high (MSI-H), microsatellite stable (MSS), or microsatellite equivocal (MSE).
- MSI testing for paired tumor-normal patients may use reads mapped to the microsatellite loci with at least five, ten, fifteen, etc. bp flanking the microsatellite region. A minimum read threshold may be used. For example, the identification of at least 10,
- mapping reads in both tumor and normal samples may be required for the locus to be included in the analysis.
- a minimum coverage threshold may be used. For example, At least 10, 15, 20, etc. of the total microsatellites on the panel may be required to reach the minimum coverage.
- Each locus may be individually tested for instability, as measured by changes in the number of nucleotide base repeats in tumor data compared to normal data, for example, using the Kolmogorov-Smirnov test. If p ⁇ 0.05, the locus may be considered unstable.
- the proportion of unstable microsatellite loci may be fed into a logistic regression classifier trained on samples from various cancer types, especially cancer types which have clinically determined MSI statuses, for example, colorectal and endometrial cohorts.
- the mean and variance for the number of repeats may be calculated for each microsatellite locus.
- a vector containing the mean and variance data may be put into a support vector machine classification algorithm. Both algorithms may return the probability of the patient being MSI-H as an output which may be compared to a threshold value.
- the sample may be classified as MSI-H. If there is between a 30-70% probability of MSI-H status, the test results may be too ambiguous to interpret and those samples may be classified as MSE. If there is a ⁇ 30% probability of MSI-HMSI-H status, the sample may be considered MSS.
- a patient report may be generated at a process 16.
- the report may be presented to a patient, physician, medical personnel, or researcher in a digital copy (for example, a JSON object, pdf file, or an image on a website or portal), a hard copy (for example, printed on paper or another tangible medium), as audio (for example, recorded or streaming audio), or in another format.
- a digital copy for example, a JSON object, pdf file, or an image on a website or portal
- a hard copy for example, printed on paper or another tangible medium
- audio for example, recorded or streaming audio
- the report may include information related to the lost or present FILA alleles and/or FILA mutations, including clinical trials for which the patient is eligible, therapies that may match the patient (for example, the systems and methods may be used as a companion diagnostic for these therapies) and/or adverse effects predicted if the patient receives a given therapy, based on the present or lost HLA alleles and/or HLA mutations in the patient’s tumor (obtained using a process 24).
- the report may include information related to whether the patient's tumor is potentially-resistant to H LA-restricted immunotherapies (for example, cellular TCR therapies, vaccines, and immunotherapies designed to be most efficacious in the presence of a particular HLA allele or alleles, etc.).
- the report may include information related to whether the patient’s tumor is potentially a good candidate for H LA-restricted immunotherapies (for example, cellular TCR therapies, vaccines, and immunotherapies designed to be most efficacious in the absence of a particular HLA allele or alleles, etc.).
- the report may state that the patient may not respond to immunotherapies that target HLA alleles that have been lost and/or mutations that are not detected in the patient sample, may or may not be eligible for clinical trials listing the loss or presence of those HLA alleles and/or mutations as inclusion or exclusion criteria (obtained using a process 26).
- treatments for example, immunotherapies
- treatments based on any HLA alleles and/or mutations present in the patient sample
- the systems and methods may be used as a companion diagnostic for these treatments
- the patient may be eligible for clinical trials listing present HLA alleles and/or mutations as inclusion criteria, and may not be eligible for clinical trials listing present HLA alleles and/or mutations as exclusion criteria (as obtained using process 26).
- the report may further include the copy number status of chr 6 and/or arms or portions of chr 6 in the tumor sample and/or normal sample.
- the report may infer HLA-LOH for that sample.
- information related to a loss of a portion of chr 6 does not specify which copy of an HLA allele was contained on the lost copy of a portion of chr 6 but provides supporting evidence that one of the HLA alleles was lost.
- the allele specific systems and methods described herein conclude that coverage of Allele B is lower than coverage of Allele A, but the coverage of Allele B is close to the threshold for calling LOH, resulting in an equivocal LOH call, which may be caused by standard variability in coverage or may reflect a partial loss or actual loss of the HLA allele.
- the chr6 LOH status serves as an orthogonal way to confirm that loss or presence of the HLA allele.
- the HLA allele that was called as equivocal loss status by the systems and methods described herein may be called as LOH.
- the HLA allele with an equivocal LOH call may be determined to be present.
- the HLA-LOH and/or HLA mutation results may be used to analyze a database of clinical data, especially to determine whether there is a trend showing that a therapy slowed cancer progression in other patients having the same or similar lost/present/mutation status as the results for a given HLA allele.
- the HLA LOH and/or mutation results may also be used to design tumor organoid experiments.
- an organoid may be genetically engineered to have the same HLA alleles and/or mutations present as a patient and may be observed after exposure to a therapy to determine whether the therapy can reduce the growth rate of the organoid, and thus may be likely to reduce the progression of cancer in the patient associated with the specimen.
- FIG. 2 illustrates an overall schematic of an example process 100 for HLA detection and analysis that may be performed by an HLA and HLA-LOH analysis system, such as that shown in FIG. 10.
- the HLA and HLA-LOH analysis system access stored genomic sequence data collected from normal tissue and from cancer tissue. More specifically, in the illustrated example, the process 100 accesses BAM files 102 containing non-cancer specimens with sequence data stored in a normal BAM file 104 and/or cancer specimens with sequence data stored in a tumor BAM file 106.
- the process 100 retrieves normal tissue (or blood) HLA mapping reads 108 from the normal BAM file 104 and tumor tissue HLA mapping reads 110 from the tumor BAM file 106.
- the normal tissue HLA mapping reads and the tumor tissue HLA mapping reads, from files 108 and 110, respectively, are communicated to or accessed by an alignment process 112.
- the alignment process 112 aligns tumor tissue data from the BAM file 106, i.e., the tumor HLA mapped reads 110, with normal tissue data from the BAM file 104, i.e., the normal HLA mapped reads 108.
- the alignment process 112 applies one or more read filters to the BAM file data prior to alignment. These filters may be applied to each HLA mapped reads data, normal tissue and tumor tissue. The filters may be applied to only one of the HLA mapped reads, normal tissue or tumor tissue.
- the filters may be stored in a hierarchical manner by the HLA and HLA-LOH analysis system, where the system applies a filters in order based on ranking, with higher ranking filters applied before lower ranked filters, and, in some examples, with an assessment of filter performance, whereby if a higher ranked filter achieves a desired filtering result, lower ranked filters are not executed by the system.
- the output from the alignment process 112 is provided to a coverage statistics process 114, that compares the aligned HLA mapped reads for normal tumor tissue and calculates coverage metrics for each allele for the normal tissue and tumor tissue data.
- the process 114 generates a report in the form of HLA allele-based coverage data 116, where that report may be stored in the system, displayed to medical personnel, and/or sent to a networked connected device, database, etc. In this way, the processes 112, 114, and 116 form an example HLA typing process.
- the HLA allele-based coverage data 116 is provided to an HLA-LOH process 118, which in the illustrated example is configured to receive other data, such as tumor purity data, tumor ploidy data, and/or genome-wide LOH predictions (collectively 120), and apply integrated metrics for performing an HLA-LOH classification on the received HLA allele-based coverage data.
- HLA-LOH process 118 which in the illustrated example is configured to receive other data, such as tumor purity data, tumor ploidy data, and/or genome-wide LOH predictions (collectively 120), and apply integrated metrics for performing an HLA-LOH classification on the received HLA allele-based coverage data.
- the process 100 includes analyzing the BAM files 102 and additionally retrieving unmapped/discarded reads (i.e., reads from a BAM file that are either assigned locations within HLA gene loci or flagged as unmapped).
- the HLA and HLA-LOH analysis system executes a preprocessing script that formats the unmapped reads (and the HLA mapped reads) from the BAM files 104 and 106 into two FASTQ files, which are fed into the next process. For the two FASTQ files, one FASTQ file is generated and contains all of the forward reads from each paired-end read, while the other FASTQ file contains the reverse of each paired-end read.
- the pairs are listed in corresponding order in the files, so the first read in the first FASTQ file will be the pair of the first read in the second FASTQ file.
- both forward and reverse reads could be included in the same FASTQ file as alternating sequences that share a similar read name.
- single read sequencing data could be included in a single FASTQ, or paired reads could be considered independent, disregarding their forward or reverse status and included in a single FASTQ.
- FIG. 3 illustrates example process 200 for the data flow for the HLA typing and the HLA-LOH model that may be implemented through the process 100.
- the two FASTQ files may be used for both FILA typing to generate FILA type, and for the LOFI model, which also receives the FILA type/patient reference as input.
- BAM files 202 are accessed on the FILA and FILA- LOFI analysis system. These BAM files 202 may be stored on the system, generated from tissue and/or blood biological samples from a subject and from populations of subjects, or generated remotely and accessed by the system, for example, through a bioinformatics pipeline that includes network accessible NGS systems or databases.
- FASTQ files 204 are generated from the BAM files 202.
- the FASTQ files 204 may include a FASTQ file that contains all of the forward reads from each paired-end read, and another FASTQ file that contains the reverse of each paired-end read.
- the FASTQ files 204 may consist of a single FASTQ file that contains single end reads, or paired end reads that are being considered as independent reads.
- the FASTQ files 204 are provided to two different processes, an FILA typing process 206 and an FILA-LOFI process 208.
- the FILA typing process 206 generates candidate alleles in the form of FILA type data 210 for the subject’s sequence data in the BAM files 202 sample.
- the FILA-LOFI process 208 generates FILA-LOFI data 212 for the subject’s sequence data.
- Each of the FILA type data 210 and the FILA-LOFI data 212 may be stored by the FILA and FILA-LOFI analysis system and reported to clinicians or other personnel.
- an alignment is performed on the sequencing data in the BAM files 202, wherein the sequencing data is aligned against a reference genome. Further, the genetic positions indicating locations in the reference genome of mapped reads having a sequence that map to the reference genome is determined. Further still, unmapped reads in the next generation sequencing data are determined, as well, and the mapped reads data and unmapped reads data are stored in one or more FASTQ files 204 having sequence reads.
- sequence read FASTQ files 204 are fed to the processes 206 and 208.
- the process 206 identifies candidate FILA alleles and stores the candidate FILA alleles as the FILA type data 210 in an FILA reference file.
- the FILA type data 210 from the process 206 is additionally fed to the FILA-LOFI process 208, which determines the FILA- LOFI status for each identified FILA allele.
- the data 210 and 212 are then stored and a report of the FILA-LOFI statuses for each of the FILA alleles may be generated.
- an HLA typing algorithm which may include the Optitype HLA Typing algorithm (Szolek et al., OptiType: precision HLA typing from next-generation sequencing data, Bioinformatics 2014, which is hereby incorporated by reference and in its entirety for all purposes) or the Kourami HLA typing algorithm (Lee et al., Kourami: graph-guided assembly for novel human leukocyte antigen allele discovery, Genome Biology 2018, which is hereby incorporated by reference and in its entirety for all purposes), may be applied to the two FASTQ files 204 input to the HLA typing process.
- the Optitype HLA Typing algorithm Szolek et al., OptiType: precision HLA typing from next-generation sequencing data, Bioinformatics 2014, which is hereby incorporated by reference and in its entirety for all purposes
- Kourami HLA typing algorithm Lee et al., Kourami: graph-guided assembly for novel human leukocyte antigen allele discovery, Genome Biology 2018, which is hereby incorporated by reference and
- the HLA typing algorithm finds mapped reads (pairs of reads) and analyzes them to predict which HLA alleles the patient has. For example, the HLA typing algorithm generates a list of predicted HLA alleles for the sample, based on reads that map to either the original reference HLA or any known HLA genetic sequence, including those in the international ImMunoGeneTics (IMGT) database. In one example, the sequences of some of the most common Class I HLA alleles are well-characterized and available to download through the IMGT (imgt.org). In one example, there are at least 40,000 known HLA genetic sequences.
- IMGT international ImMunoGeneTics
- the Optitype HLA Typing algorithm is used.
- the Optitype HLA Typing algorithm works on the premise that the correct genotype explains the source of more reads than any other genotype, where an allele is said to explain a read if the read is aligned to it with no more mismatches than to any other allele.
- the HLA Typing algorithm finds an allele combination, which maximizes the number of reads they explain.
- the HLA Typing algorithm includes three main steps. First, reads are mapped against a carefully constructed HLA allele reference. Because only exon 2 and 3 subsequences are available for all alleles, these regions are considered during read mapping so that no allele is disadvantaged because of incomplete sequence information.
- HLA typing algorithm may include flanking intronic regions and a process to impute missing sequence data based on phylogenetic information.
- a binary matrix is generated indicating which alleles a specific read could be aligned to with the least number of mismatches.
- a special case of the set cover problem is formulated as an integer linear program (ILP) that selects up to two alleles for each locus simultaneously, maximizing the number of mapped reads that can be explained by the predicted genotype.
- ILP integer linear program
- the Kourami HLA typing algorithm is a graph-guided assembly technique for classical HLA genes, which can construct allele sequences given high-coverage whole-genome sequencing data.
- the Kourami HLA typing algorithm takes advantage of partial-order graphs (POGs) to capture all known alleles.
- the Kourami HLA typing algorithm further modifies the graph to include variants found in the sequencing data so that the graph includes the paths of true alleles.
- a comprehensive reference panel is created from a combined multiple sequence alignment (MSA) of both full- length and exon-only known alleles for each HLA locus.
- MSA multiple sequence alignment
- Reads mapped to all known HLA loci in the human reference genome are extracted and aligned to the comprehensive reference panel.
- Gene-wise POGs are constructed using the combined MSAs.
- the alignments of the extracted reads are projected onto the graphs so that each read alignment is stored as a path in the graphs and the read depths on the edges naturally become edge weights.
- the graphs are modified by adding nodes and edges to incorporate differences found by the alignment, such as substitutions and indels.
- a sequence of an allele may be encoded as a path through the entire graph.
- Kourami HLA typing algorithm formulates the problem of constructing the best pair of HLA allele sequences as finding the pair of paths through the graph.
- the Kourami HLA typing algorithm considers consistent phasing information from the reads and coverage using base quality scores. Additionally, the pair of paths may be identical, to permit homozygous alleles.
- Table 1 includes 150 examples of Class I HLA alleles.
- the HLA alleles identified are HLA-A Allele 1 : A * 02:01 , HLA -A Allele 2: A * 01 :01 , HLA-B Allele 1 : B * 07:02, HLA-B Allele 2: B * 07:02, HLA-C Allele 1 : C * 07:01 , HLA-C Allele 2: C * 07:02.
- the HLA typing algorithm generates an accession number, which allows the user to retrieve an allele sequence.
- the output from the HLA typing algorithm is provided to downstream HLA-LOH models, e.g., the process 208.
- the process 100 uses the list of predicted HLA alleles, such as data 210, to create a preliminary HLA reference file composed of reference sequences of the patient’s predicted HLA alleles and all HLA pseudogenes.
- the HLA reference file is automatically generated.
- the HLA reference file may be automatically generated by pulling sequences from the Optitype (github) source code, especially the Optitype database/reference library (including the IMGT dataset) or Kourami reference library based on allele and accession number, for example using a data converter to maintain allele nomenclature consistency.
- predicted Class I HLA type data 122 is obtained and an HLA reference file is generated at a process 124, by adjusting to match the predicted HLA alleles of the non-cancer specimen.
- the process 124 generates a patient- specific HLA reference file by writing the sequence associated with each of the patient’s predicted Class I HLA types to a FASTA file.
- a FASTA file is essentially a text file where lines alternate between a sequence name (these lines start with a > symbol by convention followed by the sequence name, for example, HLA00001) and the following line is the nucleotide sequence corresponding to that sequence name.
- the process 124 writes the name and sequence for each predicted Class I HLA type as well as the pseudogenes.
- the output from the process 124 is an HLA reference file as a FASTA file that, in various embodiments, is then converted or indexed to a novoalign index file for alignment to generate a .nix file.
- the .nix file is a specialized format that allows novoalign software to more quickly and efficiently align reads. If the patient is homozygous for a given allele, it is included only once in the reference.
- This HLA reference file then may be a patient specific HLA reference file.
- the HLA reference file is a sequence file that includes the patient’s predicted HLA class I genes and all nonclassical HLA genes and HLA pseudogenes to ensure that a read maps to the correct gene, even though there is high homology from gene to gene.
- the HLA reference file is expanded to include class II HLA genes.
- a process 126 aligns HLA mapping reads, along with unmapped/discarded reads (from the two paired end FASTQ files mentioned above), to the predicted patient reference file (which is the FASTA file that has been indexed to be a .nix file), for example, using Novoalign to generate a BAM file.
- the process 126 may filter the BAM file (in one example by using pySAM) using various filtering criteria, such as, for example, checking that: (1) the read is properly paired, (2) the read is not qcjail (failed by quality control checks), (3), the read is not a duplicate, (4) the edit distance to the reference sequence of the predicted allele is less than or equal to 2, (5) the read has less than or equal to 2 insertions compared to the reference sequence of the predicted allele, (6) the read has less than or equal to 2 deletions compared to reference sequence of the predicted allele, and/or (7) both ends of paired read must map to the same predicted allele.
- a filtered BAM file is generated as a result.
- the process 126 may apply a variant calling process performed on the filtered alignment file (for example, the filtered BAM file), using freebayes (available from github), to identify any nucleotide positions where the patient’s HLA sequences diverge from the HLA reference.
- implementation of the variant calling included the following criteria: the sequence data must include at least 3 reads supporting the variant (indicating that the patient has an alternate allele, meaning a sequence that is not identical to the reference sequence of the predicted allele), and fewer than 5 reads supporting the reference sequence of the predicted allele.
- a process 128 updates the patient specific reference by replacing portions of the reference sequences with the variant sequences that are supported by at least 3 reads at the genomic positions of those variants to generate an updated patient HLA reference file.
- the updated patient HLA reference sequence file has been adjusted to match the exact nucleotide sequence of the non-cancer specimen HLA genes.
- the sequence is contained in a FASTA file that is then converted to a novoalign index file. If the patient is homozygous for a given allele, the sequence is included only once in the reference.
- the updated HLA reference file may then be sent to the process 112.
- a Novoalign alignment of HLA mapping reads is repeated along with aligning unmapped/discarded reads to the updated reference file (if updates were made).
- Strict filtering may be used, including read is properly paired; read is not qcjail; read is not a duplicate; edit distance to reference is 0; read has zero insertions to reference; read has zero deletions to reference; read is not mapped more than once.
- the process 112 aligns HLA mapping reads along with unmapped/discarded reads, to the patient HLA reference sequence (the updated HLA reference sequence data from process 128) using Novoalign and filters reads with pySAM, using strict filtering criteria to generate a cancer specimen BAM file.
- the process 114 receives the aligned HLA mapping reads and data from the process 112 and calculates coverage (for example, the number of reads that map to a single nucleotide position) for normal HLA reads.
- coverage may be inferred for nucleotide positions located between two appropriately-oriented paired reads, for example, if the two non-overlapping reads that comprise a paired-end read do not explicitly include a nucleotide position, but flank the nucleotide position, the presence of a molecule containing this intervening nucleotide position can be inferred, and thus the paired-end read may be included in the coverage metrics calculation for that nucleotide position.
- this paired-end read would count as a read that maps to the nucleotide position even though the nucleotide position is located between the two ends of the paired-end read.
- the process 114 uses bedtools to assess coverage across each of the predicted HLA alleles in the non-cancer specimen BAM file. The result is a Table of Positional Coverage across each HLA allele in the non-cancer specimen.
- the process 114 generates a csv file (116) with the number of reads that uniquely map to a specific HLA allele at each nucleotide position along that allele in the noncancer specimen.
- each column in the csv file represents a nucleotide position in an HLA gene and each row represents an allele.
- Each entry is a number representing the number of reads at that nucleotide position for that allele.
- the process 114 further calculates coverage for tumor HLA reads, e.g., using bedtools to assess coverage across each of the predicted HLA alleles in the cancer specimen BAM file.
- the result is a Table of Positional Coverage across each HLA allele in the cancer specimen, generating a csv file (116) with the number of reads that uniquely map to a specific HLA allele at each nucleotide position along that allele in the cancer specimen.
- the positional coverage for both the non-cancer and cancer specimen are contained in one csv file.
- row 1 may represent allele A in the normal sample
- row 2 may represent allele B in the normal sample
- row 3 may represent allele A in the tumor sample
- row 4 may represent allele B in the tumor sample.
- the cancer specimen is circulating tumor DNA (ctDNA) obtained from a blood sample and the coverage obtained from NGS analysis of ctDNA may differ from coverage obtained from NGS analysis of a specimen that contains solid tumor tissue or cancerous blood cells. The calculation of coverage metrics may be adjusted accordingly.
- the process 114 combines data from the Table of Positional Coverage across each HLA allele in the non-cancer specimen and the Table of Positional Coverage across each HLA allele in the cancer specimen, to generate higher level features to describe relative changes in coverage between the non-cancer specimen and cancer specimen and a Combined Coverage Metrics Table (e.g., using formulae for calculating, one example may include formulae from the following Python packages: pandas, NumPy, SciPy).
- This process 114 may generate a Combined Coverage Metrics Table, in the form of an expanded csv file that contains positional statistics on not only coverage depth but features including allelic frequencies of each allele, log ratios of each allele between tumor and normal, and areas of low sequencing coverage (See FIG. 9 for more details).
- the process 114 may also generate a Summary Statistics Table, in the form of a csv file where each row is an HLA gene and the columns contain summary statistics describing the differences in allele level coverage that will be used to make HLA LOH determinations.
- FIG. 4 illustrates an example output report displaying the results of HLA-LOH classification.
- HLA-LOH there are two detected copy losses (HLA-LOH) for HLA class I genes.
- HLA-A * 02:01 an HLA-A allele
- HLA-B * 45:02 an HLA-B allele
- No HLA-C alleles or HLA class II genes are reported lost in this example. All HLA alleles without the copy loss designation have been detected as present in the specimen.
- the report may include information related to the lost or present HLA alleles, including clinical trials for which the patient is eligible, therapies that may match the patient and/or adverse effects predicted if the patient receives a given therapy, based on the present or lost HLA alleles in the patient’s tumor.
- the report may include information related to whether the patient's tumor is potentially-resistant to H LA-restricted immunotherapies.
- the report may state that the patient may not respond to immunotherapies based on those lost HLA alleles, may not be eligible for clinical trials listing those lost HLA alleles as inclusion criteria, and may be eligible for clinical trials listing those lost HLA alleles as exclusion criteria.
- immunotherapies based on any present HLA alleles may be matched to the patient and the patient may be eligible for clinical trials listing present HLA alleles as inclusion criteria, and may not be eligible for clinical trials listing present HLA alleles as exclusion criteria.
- FIGS. 5A-5C are plots of combined coverage metrics for different examples of the techniques herein, some in comparison to non-technique examples, and some without the filter steps. (See, FIGS. 9A-9D for more details).
- FIG. 5A shows data that were calculated using all disclosed steps and features
- FIG. 5B shows data calculated without aligning discarded/unmapped reads to HLA genes
- FIG. 5C shows data calculated without replacing the HLA reference sequences with the variants detected in the sequence data generated by the patient sample.
- the process 100 may determine and report LOH Status for each HLA allele in the cancer (tumor) sample, with reference to the non-cancer (normal) sample.
- the process 118 may report all HLA alleles present in the tumor sample (known as stable alleles, versus lost alleles that are missing, absent, or detected with low coverage from the tumor sample) or, the process 118 may compare the tumor sample to a normal sample from at least one distinct patient, where that normal sample may have matched HLA types similar to the HLA types in the tumor sample to control for sequencing bias caused by hybrid capture, GC content etc.
- a selected number of normal samples from non-patients may be used by the process 118, for example, by using a panel of normal samples from nonpatients determined to have the same HLA genotype as the patient.
- normal samples may be selected so that the non-patient individuals all have the same HLA genotype that the patient does, and any normal samples that do not match the patient HLA genotype will be excluded from the normal panel for the analysis.
- the more pure a tumor sample is the stronger and more easily detectable a signal will be for a lost allele. As tumor purity decreases, the signal becomes increasingly hard to distinguish from background noise.
- the features from the Summary Statistics Table (116) are input into a machine learning classification model (of process 118) that returns a likelihood of LOH.
- alleles with a likelihood of LOH greater than 50% are reported as LOH.
- LOH Status Predictions for each allele in the predicted HLA alleles are determined by the process 118 using a Shallow Decision Tree machine learning model.
- FIG. 6 illustrates an example shallow decision tree 300 that may be executed by the process 118.
- the first line of each node (represented by a box in FIG. 6) is the name of a feature that corresponds to a statistic selected from the Summary Statistics Table (116) and a cut-off threshold against which the sample’s value for that feature is compared. If the value of the sample meets or does not meet the threshold criterion, the sample is sorted into the corresponding branch of the decision tree.
- delta_expected_difference_logR of a sample is less than or equal to 0.123
- the mean differenceJogR of the sample is then compared to a set threshold, etc.
- the other lines of text in a box may indicate the gini index value for that node, the number of samples (which may mean the number of HLA genes that were analyzed for LOH) sorted by that node, and during model training, “value” may act as a confusion matrix by indicating the number of samples (HLA genes) that were sorted into that node and that had manual annotations of either loss (right number) or stable (left number) HLA status.
- the decision tree 300 is shallow/short with few nodes to avoid overfitting, decisions are based on features from the Summary Statistics Table (116), and features or threshold values may change).
- a decision tree that is shallow may be easier to interpret, making it easier to explain the classification of a patient or specimen, for example, if a physician calls to ask about a “borderline” allele.
- the classification models of process 118 may be particularly configured to reduce processing time and increase the speed by which particular alleles can be classified, for faster ultimate diagnosis. These decision tree models are also typically more resilient to variations in upstream sample analysis.
- decision tree outputs are more discrete, for example, three possible decision tree outputs could be clear loss of an HLA allele, or clear stability of an HLA allele, and one intermediate state. Another example may include more than one intermediate state.
- LOH Status Predictions from the process 118 may be determined using other decisional techniques, such as Random Forest methods which may be slightly more accurate, and may yield a more continuous distribution of probabilities/likelihoods, for example, 75% likelihood of a loss of an HLA allele.
- the process 118 may apply a coverage threshold, such that any HLA allele with coverage below a threshold is reported by the process 118 as a loss of heterozygosity for that allele.
- the process 118 may be configured such that the threshold may be specific to the testing panel used for NGS sequencing.
- the coverage threshold below which an allele is reported as lost may be approximately 75 reads for an example (targeted -600 gene) genomic sequencing panel or 35 reads for an example (whole exome) sequencing panel, where the process reports each allele as either stable or lost.
- the model may report an equivocal or uncertain status for an allele in a specimen that is not obviously stable (present in the specimen) or lost (absent from the specimen).
- coverage metrics for an allele may fall in the middle of the distribution of coverage metrics values observed from all specimens, placing the coverage metrics in a range where the allele has a roughly equivalent probability of being either lost or stable.
- the process 100 may match a patient with clinical trials and/or a therapy/therapies that are likely to eliminate the cancer cells, based on HLA alleles that are present in cancer sample as predicted by the HLA LOH model. This may help a physician make a therapy decision or identify a matched set of possible therapies or clinical trials in which the patient may participate.
- the clinical trials are matched to the patient’s HLA LOH results based on the trials having inclusion/exclusion criteria based on the presence of specific HLA alleles in tumor or cancer cells.
- a biological assay to test for the presence of any of the alleles is performed.
- an assay which may include fluorescence activated cell sorting (FACS)
- FACS fluorescence activated cell sorting
- an assay which may include fluorescence activated cell sorting (FACS)
- FACS fluorescence activated cell sorting
- Antibodies directed to other alleles are known in the art, and additional antibodies to detect other HLA alleles are in development.
- the techniques described herein were used to analyze a patient non-cancer sample, a patient cancer sample, and a tumor organoid (T.O.) derived from the patient cancer sample and predicted that the cancer sample and T.O. had lost an A * 02 HLA allele but maintained a stable A * 03 HLA allele (see FIGS. 8A-8C).
- T.O. tumor organoid
- FACS was used on the T.O. to detect the presence of these two HLA alleles, and the results are shown in FIGS. 7A & 7B.
- FIGs. 7A & 7B include the following FACS plots: the top row shows FACS results from an anti-A * 03 antibody assay (FIG. 7A) and the bottom row shows FACS results from an anti-A * 02 antibody assay (FIG. 7B). From left to right in each row, there is a plot for a negative A * 02 control sample, a plot for the tumor organoid sample, and a plot for a positive A * 02 control. The upper half of each plot indicates which cells bound the pan HLA Class-I antibody, indicating that those cells were expressing HLA Class-I molecules.
- each plot indicates which cells bound either the anti-A * 03 antibody (top row) or the anti-A * 02 antibody (bottom row), indicating that those cells expressed the allele targeted by the antibody used to generate that plot.
- Horizontal and vertical lines within the plots indicate the location of cut-offs used to determine those percentages and numbers in the outer corners of the plots indicate the percentage of all data points in the plot that are located in each quadrant of the plot.
- Each of the plots shows a cell population that expressed HLA Class-I molecules, demonstrated by the data points being located in the upper two quadrants of each plot.
- the A * 02 negative control and the tumor organoid plots in the bottom row show a cell population that is not expressing the A * 02 allele, demonstrated by the data points being located in the left two quadrants of the plots. All remaining plots show a cell population that expressed either the A * 02 allele (bottom row plots) or the A * 03 allele (top row plots), demonstrated by the data points being located in the right two quadrants of each plot.
- T.O. tumor organoid
- T.O. genetic material may be sequenced to generate T.O. sequence data
- HLA LOH model may be used on the T.O sequence data.
- FIG. 8A-8C show examples of plots for different types of tissues. In this example, FIG.
- FIG. 8A shows coverage data calculated by the methods disclosed herein for the non-cancer sample tissue.
- FIG. 8B shows coverage data calculated by the methods disclosed herein for the cancer sample tissue.
- FIG. 8C shows coverage data calculated by the methods disclosed herein for a tumor organoid derived from the cancer sample tissue.
- FIG. 8A shows approximately equivalent coverage for two FILA alleles (A * 02:01 shown in red data points and A * 03:01 shown in blue data points) in the non-cancer tissue.
- FIG. 8B shows reduced coverage for the A * 02:01 allele.
- the sequence reads from the cancer tissue mapping to the A * 02:01 allele may be explained by the presence of non-cancer cells in the cancer sample due to the heterogeneity of cancer samples that do not have 100% tumor purity.
- FIG. 8A shows coverage data calculated by the methods disclosed herein for the non-cancer sample tissue.
- FIG. 8B shows coverage data calculated by the methods disclosed herein for the cancer sample tissue.
- FIG. 8C shows coverage
- the complete loss of the A * 02:01 allele in the T.O. may reflect the absence of non-cancer cells in the T.O., which indicates that the T.O. has 100% “tumor purity”.
- FIGS. 9A-9D illustrate example plots of coverage (number of reads) on the y-axis (plots in the top row) or the fraction of cancer specimen coverage divided by non-cancer specimen coverage (B allele fraction) on the y-axis (plots in the bottom row). These data are plotted for two FILA alleles (plotted as data points having either shades of red or shades of blue, depending on which allele is associated with each data point) at each nucleotide position indicated by the x-axis. In this example the two alleles are B * 44:03 (red data points) and
- the Optitype dataset is optimized to have consistent sequence lengths across each allele, inferring missing intronic sequence when missing, which reduces the need to normalize LOH signal across sequences of highly divergent lengths (e.g., if one allele is 1400bp and the other is only 400bp).
- the present technique first performs an alignment step using the patient’s normal NGS data allowing for some degree of mismatch. By performing variant calling against the initial HLA reference, positions where the NGS data does not support the initial chosen reference can be identified. The reference can then be updated and the alignment repeated with the more appropriate reference sequence.
- HLA-A and HLA-C are the most divergent, and yet still most alleles of these two genes share greater than 90% homology with one another across their most polymorphic regions (Exons 2 and 3). Because of this homology, including all of the patient's alleles in the mapping reference ensures that reads do not erroneously cross map between HLA genes or multimap to two HLA genes and skew coverage metrics.
- the Loss of Heterozygosity determination may hinge on whether there is a relative loss of coverage for a particular HLA allele in a tumor sample, relative to its matched normal control. This calculation may include normalizing the read counts between normal and tumor NGS data when they may have been sequenced at different depths.
- the metric used for normalization may include the number of unique reads mapping to the HLA reference, total reads, total mapped reads, or total mapped reads minus duplicates.
- Power to distinguish LOH is more of a function of coverage and estimates of tumor purity.
- these area- based metrics when integrated with depth and coverage features, also incorporate some measure of how confident the model is in its ability to resolve the two alleles (e.g. a higher area- based score means there are more positions that meet the read depth threshold and diverge between the two alleles).
- the difference in area between the variant allele frequency (VAF) curves as a feature - the B allele frequency (BAF) at any given position is the ratio of reads supporting each allele.
- the area between the two BAF curves defines how much the NGS reads have been skewed towards a particular allele.
- the BAF is almost 1.0 and 0 for the stable and lost allele, respectively.
- the tumor specific difference in BAF is an incredibly sensitive metric of allele loss.
- the BAF will fluctuate across the length of a gene but generally land somewhere around 0.5 for each allele, however it is not impossible for one allele to be slightly more well covered than the other (possibly due to better homology with sequencing probes).
- the method arrives at a feature that is robust to noise and still very sensitive to allelic imbalance.
- Tumor samples that are prepared for sequencing by NGS are generally heterogeneous and contain a mixture of tumor cells, healthy stroma and immune cells. As a result, a fully clonal loss may not necessarily appear as full loss of one allele sequence. For the sequencing specimen, it is advantageous to account for tumor purity when determining how much loss would be expected.
- Tumor purity may be estimated by methods that include but are not limited to assessing a histopathological slide corresponding to the sample that was sequenced by NGS, by analyzing DNA sequence data, or by analyzing RNA sequence data. Tumor purity may be determined manually (for example, by a trained medical professional) or by an automated, computer-assisted method.
- Expected difference in logR may be defined as log2 of (1- tumor purity). [171] Calculating delta_expected_difference_logR. An area wise difference between the observed difference in logR value and the expected difference in logR value for a complete LOH sample, defined in this patent as delta_expected_difference_logR, may be determined by comparing the observed difference in logR to the expected value generated by our tumor purity estimate, the method more effectively determines whether the loss of HLA reads observed in the tumor sample represents a loss that would be on par with clonal LOH.
- a loss of heterozygosity in a specific HLA gene (such as HLA-A, HLA-B, or HLA- C) in a cancer specimen may be determined in accordance with a threshold value, which may be set if, for instance, a significant difference exists between the read counts of the first tumor allele for the HLA gene and the read counts of the second tumor allele for the HLA gene.
- a significant difference may exist, for instance, if the difference between the read counts of the first tumor allele for the HLA gene and the read counts of the second tumor allele for the HLA gene is significantly more than the difference between the read counts of the first normal allele for such HLA gene and the read counts of the second normal allele for such HLA gene.
- “Significantly more” may be confirmed, for instance, when the delta_expected_difference_logR value for the HLA gene is significant. For instance, the delta_expected_difference_logR value may be significant if it is between 0 and -2. “Significantly” more may be confirmed, for instance in circumstances where LOH is partial rather than complete, when the delta_expected_difference_logR value for the HLA gene is between 0 and .1 , between 0 and 0.2, between 0 and 0.25, between 0 and 0.5, or between 0 and 1.
- Determination of whether an HLA gene suffers a LOH can help further determine whether certain treatment options may be appropriate for patients.
- treating the cancer by administering a therapy known to be effective against HLA-heterozygous cancers may be appropriate.
- a checkpoint inhibitor therapy may be appropriate for a subject with an HLA-heterozygous cancer.
- the checkpoint inhibitor therapy may be selected from the group consisting of an anti-CTLA-4 therapy, an anti-PD-1 therapy, or an anti-PD-L1 therapy, for example.
- Examples may include ipilimumab, nivolumab, pembrolizumab, pidilizumab, atezolizumab, Ipilimumab, and/or tremelimumab, and may include combination therapies, such as nivolumab + ipilimumab.
- a cancer vaccine may be appropriate, such as a cancer vaccine targeted to a specific HLA allele.
- a peptide cancer vaccine available through Shiga University to treat HLA-A * 02-positive advanced non-small cell lung cancer (NCT01069640).
- Another example is a peptide cancer vaccine available through Shiga University to treat HLA-A * 24-positive advanced small cell lung cancer (NCT01069653).
- FIG. 10 illustrates an example system 400 for HLA and HLA-LOH analysis that may be implemented on a network accessible processing system for performing the processes described herein.
- the system 400 may be part of a precision medicine platform.
- the example system may be part of an NGS system or implemented on one or more network accessible processing systems (e.g., servers) communicatively coupled to an NGS system, a network accessible sequencing database, digital reporting system, or other processing system.
- network accessible processing systems e.g., servers
- the HLA and HLA-LOH analysis system 400 may be configured for performing the methods described herein including those of processes 100 and 200.
- the system 400 may include a computing device 402, and more particularly may be implemented on one or more processing units 404, e.g., Central Processing Units (CPUs), and/or on one or more or Graphical Processing Units (GPUs) 406, including clusters of CPUs and/or GPUs.
- CPUs Central Processing Units
- GPUs Graphical Processing Units
- Features and functions described may be stored on and implemented from one or more non-transitory computer-readable media 408 of the computing device.
- the computer-readable media 408 may include, for example, an operating system 410 and software modules, or "engines,” that implement the methods described herein, including those of processes 100 and 200 and other processes illustrated and described herein.
- the computer-readable media 408 stores an HLA analysis system 412 for performing the HLA typing processes and HLA-LOH processes described herein.
- the HLA analysis system 412 includes an HLA typing process 414 and an HLA-LOH process 416, both similar to those described in examples of FIGS. 2 and 3.
- An HLA- LOH report generator 418 is configured to store and generate HLA allele predictions and LOH allele reports, also in accordance with the examples herein.
- the computer-readable media 408 may store sequence data processing instructions, including BAM file analysis instructions, sequence data filtering instructions, FASTQ file generation instructions, and normalization processes instructions for implementing the techniques herein.
- the computing device 402 may be a distributed computing system, such as an Amazon Web Services cloud computing solution.
- the computing device 402 may be implemented on one network accessible processing device 450 or distributed across multiple such devices 450, 452, 454, etc.
- the computing device 402 includes a network interface 420 communicatively coupled to network 422, for communicating to and/or from a portable personal computer, smart phone, electronic document, tablet, and/or desktop personal computer, or other computing devices for communicating overlay maps, predicted tile classifications and locations, predicted cell classifications and locations, etc. Such information may also be stored in a database 424.
- the computing device 402 further includes an I/O interface 426 connected to devices, such as digital displays 428 for displaying generator overlay maps, user input devices 430, etc.
- a dashboard generator 432 may be used to generate GUI and/or other digital displays allowing a user to review and interact with and adjust generated HLA allele reports and HLA-LOH allele reports.
- the network 422 may be a public network such as the Internet, a private network such as that of a research institution or a corporation, or any combination thereof.
- Networks can include, local area network (LAN), wide area network (WAN), cellular, satellite, or other network infrastructure, whether wireless or wired.
- the networks can utilize communications protocols, including packet-based and/or datagram-based protocols such as Internet protocol (IP), transmission control protocol (TCP), user datagram protocol (UDP), or other types of protocols.
- IP Internet protocol
- TCP transmission control protocol
- UDP user datagram protocol
- the networks can include a number of devices that facilitate network communications and/or form a hardware basis for the networks, such as switches, routers, gateways, access points (such as a wireless access point as shown), firewalls, base stations, repeaters, backbone devices, etc.
- the computer-readable media 408 may include executable computer-readable code stored thereon for programming a computer (e.g., comprising a processor(s) and GPU(s)) to the techniques herein.
- Examples of such computer-readable storage media include a hard disk, a CD-ROM, digital versatile disks (DVDs), an optical storage device, a magnetic storage device, a ROM (Read Only Memory), a PROM (Programmable Read Only Memory), an EPROM (Erasable Programmable Read Only Memory), an EEPROM (Electrically Erasable Programmable Read Only Memory) and a Flash memory.
- the processing units of the computing device may represent a CPU-type processing unit, a GPU-type processing unit, a field-programmable gate array (FPGA), another class of digital signal processor (DSP), or other hardware logic components that can be driven by a CPU.
- FPGA field-programmable gate array
- DSP digital signal processor
- EXAMPLE 1 [182] Methods: A total of 434 colorectal or non-small cell lung cancer samples underwent DNA sequencing on a genomic sequencing panel using paired, FFPE tumor and normal (blood or saliva) samples. To detect HLA-LOH from NGS data, we took advantage of accurate NGS-based HLA typing to resolve the patient’s most likely HLA haplotype. Based on this haplotype, we adaptively realigned reads, extracted a number of features that describe the relative allele coverage in the tumor and normal sample, and used these features to make a confident determination of allelic loss in the patient’s tumor sample.
- HLA-LOH To investigate the prevalence of HLA-LOH, we utilized the specialized pipeline described above to detect HLA-LOH by DNA next-generation sequencing (NGS).
- NGS DNA next-generation sequencing
- Class I HLA alleles are highly polymorphic and most individuals have two distinct alleles for each HLA gene. Each allele allows for presentation of a unique pool of short peptides (approximately 8-11 amino acids in length) derived from the cellular products being made by each cell in the body.
- NGS next-generation sequencing
- HLA Loss of Heterozygosity is a potential escape mechanism for tumors under immune pressure, where tumors can lose one copy of HLA and thereby avoid presenting potent neoepitopes. (See FIG. 11 and Tran et al., New England Journal of Medicine 2016;
- HLA LOH could be an especially important escape mechanism to identify in target populations.
- the HLA-LOH process 100 was used.
- the HLA- LOH process 100 takes as inputs BAM files 102 from a matched Tumor and Normal Sample, respectively, as well as two digit HLA type 122 (similar to those generated by Optitype/Kourami/etc.), and tumor purity and ploidy information 120. (See FIG. 2) A full length HLA sequence is not required.
- the process 100 then maps all HLA mapping reads as well as all unmapped reads to a new HLA reference 124 & 126. After accounting for potential germline variants present in the sample’s HLA genes, it updates alignments and determines allele specific coverage.
- the process 100 determines, at 128, whether any reduction in allele coverage is consistent with a clonal loss of a specific HLA allele.
- the output of the HLA-LOH process 100 is a prediction of LOH status for HLA-A, HLA-B, and HLA-C genes.
- HLA LOH occurs across the entire locus -
- HLA-A, HLA-B, and/or HLA-C genes One feature that stood out was the fact that in the majority of cases (44/80), when LOH was observed at one gene in the HLA locus it was also observed across the other genes in that locus (HLA-A, HLA-B, and/or HLA-C genes), suggesting that the Class I locus is often lost together (See FIG. 14).
- FIGS. 7A and 7B are flow cytometry experiment results showing the expression of the stable and lost allele relative to a pan HLA antibody. Gated on live cells.
- a machine learning algorithm MLA
- NN neural network
- MLAs include supervised algorithms (such as algorithms where the features/classifications in the data set are annotated) using linear regression, logistic regression, decision trees, classification and regression trees, Naive Bayes, nearest neighbor clustering; unsupervised algorithms (such as algorithms where no features/classification in the data set are annotated) using Apriori, means clustering, principal component analysis, random forest, adaptive boosting; and semi-supervised algorithms (such as algorithms where certain features/classifications in the data set are annotated) using generative approach (such as mixture of Gaussian distributions, mixture of multinomial distributions, hidden Markov models), low density separation, graph- based approaches (such as mincut, harmonic function, manifold regularization), heuristic approaches, or support vector machines.
- supervised algorithms such as algorithms where the features/classifications in the data set are annotated
- Apriori means clustering, principal component analysis, random forest, adaptive boosting
- semi-supervised algorithms such as algorithms where certain features/classifications in the data set are annotated
- NNs include conditional random fields, convolutional neural networks, attention based neural networks, long short term memory networks, or other neural models where the training data set includes a plurality of samples and RNA expression data for each sample. While MLA and neural networks identify distinct approaches to machine learning, the terms may be used interchangeably herein. Thus, a mention of MLA may include a corresponding NN or a mention of NN may include a corresponding MLA.
- Training may include identifying common expression characteristics shared across RNA gene expressions in tissue normal samples, primary samples, and metastatic samples, such that the MLA may predict the ratio of a metastases tumor from the background tissue and identify which portion of an input RNA expression set may be attributed to the tumor and which portion may be attributed to the background tissue.
- Common expression characteristics may include which genes are expected to be overexpressed, expressed, and/or underexpressed for each type of tissue and/or tumor and may be identified for each k cluster as the corresponding genes.
- the annotations provided for each sample would be a full transcriptome gene expression dataset, cancer type, tissue site, and background tissue percentage.
- an implementation of one or more embodiments of the methods and systems as described above may include microservices constituting a digital and laboratory health care platform supporting detection of LOH in a cancer specimen, especially in HLA genes.
- Embodiments may include a single microservice for executing and delivering HLA LOH detection or may include a plurality of microservices each having a particular role which together implement one or more of the embodiments above.
- a first microservice may execute alignment of reads to HLA genes in order to deliver HLA reference sequences to a second microservice for calculating coverage metrics.
- the second microservice may execute calculating coverage metrics to deliver coverage metrics according to an embodiment, above.
- a third microservice may receive coverage metrics from a second microservice and may execute HLA LOH modeling to deliver an LOH status for each HLA allele in a specimen.
- micro-services may be part of an order management system that orchestrates the sequence of events as needed at the appropriate time and in the appropriate order necessary to instantiate embodiments above.
- a micro-services based order management system is disclosed, for example, in U.S. Patent Application No. 16/927,976, titled “Adaptive Order Fulfillment and Tracking Methods and Systems”, filed 7/13/2020, which is incorporated herein by reference and in its entirety for all purposes.
- an order management system may notify the first microservice that an order for HLA typing has been received and is ready for processing.
- the first microservice may execute and notify the order management system once the delivery of HLA typing is ready for the second microservice.
- the order management system may identify that execution parameters (prerequisites) for the second microservice are satisfied, including that the first microservice has completed, and notify the second microservice that it may continue processing the order to calculate coverage metrics according to an embodiment, above.
- the genetic analyzer system may include targeted panels and/or sequencing probes.
- a targeted panel is disclosed, for example, in U.S. Prov. Patent Application No. 62/902,950, titled “System and Method for Expanding Clinical Options for Cancer Patients using Integrated Genomic Profiling”, and filed 9/19/19, which is incorporated herein by reference and in its entirety for all purposes.
- An example of a targeted panel for sequencing cell-free (cf) DNA and determining various characteristics of a specimen based on the sequencing is disclosed, for example, in U.S. Patent Application No.
- targeted panels may enable the delivery of next generation sequencing results for HLA LOH detection according to an embodiment, above.
- An example of the design of next-generation sequencing probes is disclosed, for example, in U.S. Patent Application No. 17/076,704, titled “Systems and Methods for Next Generation Sequencing Uniform Probe Design”, and filed 10/21/20 and U.S. Patent Application No. 17/323,986, titled “Systems and Methods for Next Generation Sequencing Uniform Probe Design”, and filed 5/18/21, which are incorporated herein by reference and in their entirety for all purposes.
- the digital and laboratory health care platform further includes an epigenetic analyzer system
- the epigenetic analyzer system may analyze specimens to determine their epigenetic characteristics and may further use that information for monitoring a patient over time.
- An example of an epigenetic analyzer system is disclosed, for example, in U.S. Patent Application No. 17/352,231, titled “Molecular Response And Progression Detection From Circulating Cell Free DNA”, and filed 6/18/21 , which is incorporated herein by reference and in its entirety for all purposes.
- the methods and systems described above may be utilized after completion or substantial completion of the systems and methods utilized in the bioinformatics pipeline.
- the bioinformatics pipeline may receive next-generation genetic sequencing results and return a set of binary files, such as one or more BAM files, reflecting DNA and/or RNA read counts aligned to a reference genome.
- the methods and systems described above may be utilized, for example, to ingest the DNA and/or RNA read counts and produce HLA LOH detection as a result.
- any RNA read counts may be normalized before processing embodiments as described above.
- RNA data normalizer An example of an RNA data normalizer is disclosed, for example, in U.S. Patent Application No. 16/581 ,706, titled “Methods of Normalizing and Correcting RNA Expression Data”, and filed 9/24/19, which is incorporated herein by reference and in its entirety for all purposes.
- any system and method for deconvolving may be utilized for analyzing genetic data associated with a specimen having two or more biological components to determine the contribution of each component to the genetic data and/or determine what genetic data would be associated with any component of the specimen if it were purified.
- An example of a genetic data deconvolver is disclosed, for example, in U.S. Patent Application No. 16/732,229 and PCT19/69161 , both titled “Transcriptome Deconvolution of Metastatic Tissue Samples”, and filed 12/31/19; and U.S. Patent Application No. 17/074,984, titled “Calculating Cell-type RNA Profiles for Diagnosis and Treatment”, and filed 10/20/20, all of which are incorporated herein by reference and in their entirety for all purposes.
- RNA expression levels may be adjusted to be expressed as a value relative to a reference expression level. Furthermore, multiple RNA expression data sets may be adjusted, prepared, and/or combined for analysis and may be adjusted to avoid artifacts caused when the data sets have differences because they have not been generated by using the same methods, equipment, and/or reagents.
- RNA expression levels associated with multiple samples may be compared to determine whether an artifact is causing anomalies in the data.
- An example of an automated RNA expression caller is disclosed, for example, in U.S. Patent Application No. 17/112,877, titled “Systems and Methods for Automating RNA Expression Calls in a Cancer Prediction Pipeline”, and filed 12/4/20, which is incorporated herein by reference and in its entirety for all purposes.
- the digital and laboratory health care platform may further include one or more insight engines to deliver information, characteristics, or determinations related to a disease state that may be based on genetic and/or clinical data associated with a patient and/or specimen.
- exemplary insight engines may include a tumor of unknown origin engine, a tumor mutational burden engine, a PD-L1 status engine, a homologous recombination deficiency engine, a cellular pathway activation report engine, an immune infiltration engine, a microsatellite instability engine, a pathogen infection status engine, and so forth.
- An example tumor of unknown origin engine is disclosed, for example, in U.S. Patent Application No.
- TMB tumor mutational burden
- PD-L1 status engine is disclosed, for example, in U.S. Patent Application No.
- 16/789,363 titled “An Integrative Machine-Learning Framework to Predict Homologous Recombination Deficiency”, and filed 2/12/20, which is incorporated herein by reference and in its entirety for all purposes.
- An example of a cellular pathway activation report engine is disclosed, for example, in U.S. Patent Application No. 16/994,315, titled “Systems And Methods For Detecting Cellular Pathway Dysregulation In Cancer Specimens”, and filed 8/14/20, which is incorporated herein by reference and in its entirety for all purposes.
- An example of an immune infiltration engine is disclosed, for example, in U.S. Patent Application No.
- Another example of a pathogen infection status engine is disclosed, for example, in PCT/US21/18619, titled “Systems And Methods For Detecting Viral DNA From Sequencing”, and filed February 18, 2021 , which is incorporated herein by reference and in its entirety for all purposes.
- An example of a T cell receptor or B cell receptor profiling engine is disclosed, for example, in U.S. Patent Application No. 17/302,030, titled “TCR/BCR Profiling”, and filed April 21 , 2021 , which is incorporated herein by reference and in its entirety for all purposes.
- An example of a line of therapy engine is disclosed, for example, in U.S. Patent Application No. 17/001,673, titled “Unsupervised Learning And Prediction Of Lines Of Therapy From High-Dimensional Longitudinal Medications Data”, and filed August 24, 2020, which is incorporated herein by reference and in its entirety for all purposes.
- the methods and systems described above may be utilized to create a summary report of a patient’s genetic profile and the results of one or more insight engines for presentation to a physician.
- the report may provide to the physician information about the extent to which the specimen that was sequenced contained tumor or normal tissue from a first organ, a second organ, a third organ, and so forth.
- the report may provide a genetic profile for each of the tissue types, tumors, or organs in the specimen.
- the genetic profile may represent genetic sequences present in the tissue type, tumor, or organ and may include variants, expression levels, information about gene products, or other information that could be derived from genetic analysis of a tissue, tumor, or organ.
- the report may include therapies and/or clinical trials matched based on a portion or all of the genetic profile or insight engine findings and summaries.
- the clinical trials may be matched according to the systems and methods disclosed in U.S. Patent Application No. 16/889,779, titled “Systems and Methods of Clinical Trial Evaluation”, filed 6/1/2020, which is incorporated herein by reference and in its entirety for all purposes.
- the report may include a comparison of the results (for example, molecular and/or clinical patient data) to a database of results from many specimens.
- results for example, molecular and/or clinical patient data
- results for example, molecular and/or clinical patient data
- U.S. An example of methods and systems for comparing results to a database of results are disclosed in U.S.
- the information may be used, sometimes in conjunction with similar information from additional specimens and/or clinical response information, to discover biomarkers or design a clinical trial.
- Any data generated by the systems and methods and/or the digital and laboratory health care platform may be downloaded by the user.
- the data may be downloaded as a CSV file comprising clinical and/or molecular data associated with tests, data structuring, and/or other services ordered by the user. In various embodiments, this may be accomplished by aggregating clinical data in a system backend, and making it available via a portal.
- This data may include not only variants and RNA expression data, but also data associated with immunotherapy markers such as MSI and TMB, as well as RNA fusions.
- the digital and laboratory health care platform further includes a device comprising a microphone and speaker for receiving audible queries or instructions from a user and delivering answers or other information
- a device comprising a microphone and speaker for receiving audible queries or instructions from a user and delivering answers or other information
- the methods and systems described above may be utilized to add data to a database the device can access.
- An example of such a device is disclosed, for example, in U.S. Patent Application No. 16/852,138, titled “Collaborative Artificial Intelligence Method And System”, and filed April 17, 2020, which is incorporated herein by reference and in its entirety for all purposes.
- the digital and laboratory health care platform further includes a mobile application for ingesting patient records, including genomic sequencing records and/or results even if they were not generated by the same digital and laboratory health care platform, the methods and systems described above may be utilized to receive ingested patient records.
- a mobile application for ingesting patient records, including genomic sequencing records and/or results even if they were not generated by the same digital and laboratory health care platform
- An example of such a mobile application is disclosed, for example, in U.S. Patent Application No. 16/289,027, titled “Mobile Supplementation, Extraction, And Analysis Of Health Records”, and filed February 28, 2019, which is incorporated herein by reference and in its entirety for all purposes.
- Another example of such a mobile application is disclosed, for example, in U.S.
- Another example of such a mobile application is disclosed, for example, in U.S. Patent Application No. 17/157,974, titled “Mobile Supplementation, Extraction, And Analysis Of Health Records”, and filed January 25, 2021, which is incorporated herein by reference and in its entirety for all purposes.
- the methods and systems may be used to further evaluate genetic sequencing data derived from an organoid to provide information about the extent to which the organoid that was sequenced contained a first cell type, a second cell type, a third cell type, and so forth.
- the report may provide a genetic profile for each of the cell types in the specimen.
- the genetic profile may represent genetic sequences present in a given cell type and may include variants, expression levels, information about gene products, or other information that could be derived from genetic analysis of a cell.
- the report may include therapies matched based on a portion or all of the deconvolved information.
- organoids may be cultured and tested according to the systems and methods disclosed in U.S. Patent Application No.
- the drug sensitivity assays may be especially informative if the systems and methods return results that match with a variety of therapies, or multiple results (for example, multiple equally or similarly likely cancer types or tumor origins), each matching with at least one therapy.
- the digital and laboratory health care platform further includes application of one or more of the above in combination with or as part of a medical device or a laboratory developed test that is generally targeted to medical care and research, such laboratory developed test or medical device results may be enhanced and personalized through the use of artificial intelligence.
- An example of laboratory developed tests, especially those that may be enhanced by artificial intelligence, is disclosed, for example, in U.S. Provisional Patent Application No. 62/924,515, titled “Artificial Intelligence Assisted Precision Medicine Enhancements to Standardized Laboratory Diagnostic Testing”, and filed 10/22/19, which is incorporated herein by reference and in its entirety for all purposes.
- techniques herein may be extended to include LOH loss type classifications, using a two-layer HLA LOH classifier model developed to classify specimens as having a LOH status of loss, no loss, and for specimens having a loss of heterozygosity a further classification of whether the loss is a complete (clonal) loss of heterozygosity (for example, nearly all of the cancer cells in a specimen are predicted to have LOH) or partial loss of heterozygosity (for example, only a portion of the cancer cells in a specimen are predicted to have LOH).
- the result is a three class model of HLA LOH status.
- these techniques can be agnostic to the type of initial HLA LOH classifier used.
- these techniques can be implemented with input classification data from the HLA LOH classifications described above with reference to FIGS. 2 and 3 and/or input classification data developed using other types of HLA LOH classifiers.
- these techniques can advantageously identify mis-classified HLA LOH status and correctly reclassify, e.g., detecting previously classified no loss of heterozygosity specimens as partial loss of heterozygosity thereby allowing for more accurate classification results that lead to more accurate decisions on matched therapy types and/or determination of meeting eligibility criteria in order to match clinical trials to a patient or specimen.
- FIG. 16 illustrates an example process 500 for determining LOH status, as may be executed by the bioinformatics pipeline 14, the computing device 400, and in particular the HLA analysis system 412.
- Biological sample data 502 is provided to HLA LOH classification process 504 that identifies the sample as corresponding to one of two LOH classifications, loss of heterozygosity or no loss of heterozygosity.
- the data 502 may include data in accordance with examples herein, such as HLA reads, HLA reference data, alignment data between the two, coverage feature metrics (e.g., statistics), and/or the determination of allelic imbalance data, etc.
- the process 504 may be implemented by the LOH modeling processes of processes 100 and 200 of FIGS.
- process 118 determines LOH status for an entire sample and/or for each HLA allele in the sample, by referencing to a normal sample.
- the classification process 504 generates one of two classification outputs that may be reported out, a “No loss” (no loss of heterozygosity (no LOH)) classification 506 or a “Loss” (loss of heterozygosity (LOH)) classification 508.
- Such determination corresponds to a first layer 510 of the two layer configuration of the process 500.
- HLA reads, HLA reference data, alignment data between the two, coverage feature metrics (e.g., statistics), and/or the determination of allelic imbalance data, etc. may be provided to a second layer 512 that contains a second HLA LOH classification process 514 designed to classify the sample as a partial LOH 516 or clonal LOH 518.
- the first layer 510 may be implemented in accordance with the example methods like that described above in reference to FIGS. 1-3. However, an advantage of some embodiments herein is that the second layer 512 may be implemented agnostic to the source of initial 2-state LOH classification data provided thereto.
- FIG. 17 illustrates an example process 600 that may be implemented by the processes 504 and 514.
- Initial HLA reads are identified or otherwise obtained, at a block 602.
- a patient-specific HLA reference (genome or partial genome) is generated using normal HLA mapping reads, at a block 604, from which the HLA reads are aligned to the HLA reference, at a block 606.
- Coverage feature metrics from that alignment are computed at a block 608 and a determination of allelic imbalance(s) are made at a block 610, from which a determination of LOH status is performed at block 612.
- the block 612 may be implemented as a two-stage LOH classification process, executing processes 504 and 514, e.g., using models based on any of the coverage features and other metrics of processes 504 and 514.
- the HLA coverage feature metrics described herein may be initially received at the process 608, without needing to be determined.
- the process 600 may be truncated to start at process 608 and the receipt of HLA coverage feature metrics that have been determined from an external source based on any number of sequence read alignment and filtering processes.
- HLA-LOH for the HLA-A gene
- HLA-A * 02:01 allele is discussed to illustrate analysis for LOH of a particular HLA allele.
- this same process applies to determining HLA-LOH for any HLA gene (i.e., HLA-A, HLA-B, HLA-C) or allele of any of those genes.
- identifying HLA reads at block 602 was performed as follows. Because a number of informative HLA reads lack sufficient homology to the standard human reference genome to successfully map during routine analysis, specimen sequence reads of interest for HLA LOH determination were collected from two sources: reads already mapped to an HLA gene in the hg19-aligned BAM output; and unmapped reads from the BAM output that align to a reference file having a large number of HLA alleles collected from a database, such as the IEDB Database. These reads are then combined together into a single file for each of the normal sample and the tumor sample and referred to as normal HLA mapping reads and tumor HLA mapping reads, respectively. Reads from the received sample are compared against these combined reads to identify the HLA reads for analysis.
- the block 604 in order to assess relative coverage across a patient’s two HLA-A alleles, and ultimately determine whether one has been lost based on HLA-A coverage in a matched tumor sample, the block 604 first identifies the sequences of the two HLA-A alleles that are present. In an example, this may be achieved using Optitype, in accordance with examples described hereinabove.
- normal HLA mapping reads are passed into Optitype and the pair of HLA sequences that explain the greatest proportion of HLA mapping reads with the least amount of error is returned (or a single sequence in the case of a homozygous sample).
- each reference allele contains intronl , exon 2, intron 2, exon 3, and intron 3 of the allele, referred to as the HLA Region of Interest.
- the reference file generated includes the sequences that were determined for the HLA-A, HLA-B, and HLA-C genes as well as a pool of non-classical HLA genes and HLA pseudogenes to minimize issues that may arise from homology between these genes and HLA-A.
- HLA-A mapping reads are re-aligned to the HLA-A specific reference generated at block 604, e.g., using a Novoalign process.
- Novoalign may be executed with parameters that allow a read to be mapped in more than one location provided those locations both have equivalent mapping qualities (as opposed to one location being selected at random).
- reads may then be removed that have more than one mismatch, insertion, or deletion relative to the reference.
- the HLA reference file may be assessed to determine whether the sequences present are fully supported by the reads in the sample. For example, Freebayes (v1.1.0) may be used to detect any positions in the HLA reference file where another germline sequence is more supported by the sequencing results.
- reads may be updated in cases where there are at least 40 reads covering a position, and fewer than 5 of those reads support the current reference position. In any case, such information may be provided to the block 604 for updating the HLA reference file or the HLA reference file may be updated at the block 606. In cases where reference updates are needed, the alignment and post alignment filtering described above is repeated at the block 606.
- computing allele coverage feature metrics and normalization may be performed as follows. Following alignment and filtering at block 606, in an example, Bedtools (v 2.26.0) was used to calculate the number of reads that support each allele at each position across a region of interest. In some embodiments, the block 608 can further perform de-noising across the region of interest. For example, to minimize the effect of fluctuations in coverage, in an example, the block 608 is configured to apply a fourth order Savitzky Golay filter with a window length set in base pairs (e.g., 801 bp) to all coverage values.
- base pairs e.g. 801 bp
- the coverage depth of each allele along the length of the region of interest is then used by the block 608 to generate a number of higher order coverage feature metrics.
- these higher order features include B allele frequency (BAF), the proportion of reads supporting each allele at each position.
- the features further include log ratio (logR), i.e., the ratio of coverage for an allele between the tumor and normal sample. A negative log ratio indicates that the allele is less abundant in the tumor than in the normal. This ratio may be calculated as the log2(tumor_read_depth/normal_read_depth * normalization factor), where the normalization factor is the ratio of the number of mapped paired primary reads in the final normal and tumor alignment files.
- the allele with the lower mean logR value is designated the “target allele”, and a further determination procedure is used to determine whether the target allele has undergone Loss of Heterozygosity relative to the “stable” allele.
- each of the processes 504 and 514 may contain trained classifier models.
- the process 504 for example, may be a classifier trained to determine allelic imbalance from which all samples with partial or clonal loss of heterozygosity are collectively classified separate from samples classified as no loss of heterozygosity.
- the process 514 may be a classifier trained for sequential assessment using specific data for tumor and normal samples with the application of predetermined thresholds. The coverage features and thresholds applied by each of the processes 504 and 514 may be established empirically using training data.
- the processes 504 and 514 may be implemented by respective logistic regression models.
- coverage feature selection and threshold determination for the models of 504 and 514 were established empirically using a training set of 189 samples across 34 cancer types that underwent manual classification by two expert reviewers to annotate partial and clonal LOH.
- 477 loci (295 with no loss, 92 with a partial loss and 90 with a clonal loss) across 186 samples with concordant results and tumor purity >30% were selected for training.
- Initial performance of both models was evaluated on a hold-out (validation) dataset of 203 loci across 77 samples (128 with no loss, 37 with a partial loss, 37 with a clonal loss).
- the process 504 performs classification for allelic imbalance based on three coverage features metrics, where training data corresponding to each feature individually and/or collectively is issued for training the allelic imbalance model at process 504.
- the first feature is the ratio of B allele frequency (BAF) of stable allele between tumor and normal samples, which captures the magnitude of the LOH signal between a tumor sample and a normal sample. While the BAF in the normal sample should always be approximately 0.5, in samples with clonal LOH, the BAF of the retained allele in the tumor will increase substantially as tumor purity increases, for example. Training data that includes normal samples and tumor samples of different BAF values, in particular at different tumor purity levels therefore may be used during model training.
- BAF B allele frequency
- the second feature is the mean difference in LogR values between the target allele and the stable alleles.
- This logR value represents the change in coverage for a given allele from the normal to the tumor sample. For an allele that has undergone LOH, its logR will decrease significantly, and the logR of the corresponding stable allele will generally increase slightly. The difference between these two values represents the magnitude of the total change.
- Training data that includes normal samples and tumor samples of alleles with different coverage amounts may be used during model training.
- the third feature is tumor purity.
- training data may include samples of different tumor purity as determined by a pathologist. As tumor purity approaches the limit of detection there may be a greater degree of uncertainty around the determination of allelic imbalance.
- the allelic imbalance model of the process 504 returns a probability of having allelic imbalance.
- samples with a probability of less than 0.5 are classified as LOH negative (classification 506); and samples that have a probability of greater than 0.5 are classified as LOH positive (classification 508) and assessed by a second classification model at process 514.
- these probability thresholds are determined by the model training and therefore may be different than that listed. Further, the model training process may determine that the cut off probabilities for determining classifications 506 and 508 may be different values.
- the process 514 uses a LOH modeling process independent from that of process 504, namely a clonal LOH model that uses three coverage features metrics to classify the loss of heterozygosity as either clonal or partial LOH.
- the coverage features of process 514 may include the expression: (observed log Ft difference - expected log Ft difference) / expected logFt difference, i.e., the ratio of the difference between observed and expected logFt difference to the expected logFt difference.
- the difference in LogFt values between the target and stable allele represent the magnitude of the loss event.
- the expected LogFt difference can be calculated as the log 2 (1-TP), where TP is the tumor purity.
- the ratio of observed to expected logFt describes whether the loss event observed meets or exceeds the expected loss for a clonal LOH event.
- the observed logFt difference is the difference between the logFt of coverage of the stable allele and the logFt of coverage of the lost allele.
- the observed logFt difference is an average of log(coverage in tumor / coverage in normal), calculated for at least one nucleotide position in an HLA gene.
- the log(coverage in tumor / coverage in normal) may be calculated for nucleotide positions having a coverage of at least 40 sequence reads.
- the observed logFt difference is an average of log(coverage in tumor / coverage in normal * match ratio), calculated for at least one nucleotide position in an HLA gene, wherein the match ratio is the ratio of the number of HLA reads in the normal sample to the number of HLA reads in the tumor sample or the ratio of the number of unique reads in the normal sample to the number of unique reads in the tumor sample.
- the log(coverage in tumor / coverage in normal * match ratio) may be calculated for nucleotide positions having a coverage of at least 40 sequence reads.
- the observed logR difference is the cumulative area between the logR line associated with a first allele and the logR line associated with a second allele.
- the expected logR difference is the log2(1 -tumor purity) and tumor purity is a value between 0 and 1.
- the process 514 may calculate, for each allele, a ratio of a BAF of a lost allele in the tumor sample to the BAF of the lost allele in the normal sample. Then, the process 514 may compare each ratio for the alleles and select the allele associated with the lowest ratio as the allele that is more likely to be lost. The process 514 may do this determination before determining LOFI classification.
- the coverage features of the process 514 may further include the ratio of BAF of the target allele between tumor and normal samples. This feature captures the magnitude of the LOFI signal between the tumor and normal sample. While the BAF in the normal sample should always be approximately 0.5, the BAF in the tumor will decrease substantially in samples with Clonal LOFI as tumor purity increases. Additionally, the coverage features of the process 514 may include tumor purity, where, as the tumor purity approaches the limit of detection, there may be a greater degree of uncertainty around the determined classification.
- the clonal LOFI model returns a probability of clonal loss of heterozygosity. If the probability of clonal LOFI is greater than 0.5, the process 514 will return a result of clonal LOFI for the target allele. For example, if the target allele is A * 02:01 , then the process 514 will return a status of “A * 02:01 LOFI Positive”, corresponding to classification 518. If the probability is 0.5 or less, the process 514 will return a result of partial LOFI for the target allele, e.g., “A * 02:01 LOFI Partial”, corresponding to classification 516.
- FIG. 18 illustrates three plots.
- a top plot is of the read coverage (number of reads) on the y-axis for two different alleles (B * 44:02 (red data points) and B * 07:02 (blue data points)) as a function of nucleotide position, for a normal sample.
- a middle plot is of the BAF for two different alleles as a function of nucleotide position, for the normal sample.
- a bottom plot is of Log Ratio of read coverage in the tumor sample to the read coverage in the normal sample, as a function of nucleotide position.
- FIG. 19 illustrates three plots.
- a top plot is of the read coverage (number of reads) on the y-axis for two different alleles (B * 44:02 (red data points) and B * 07:02 (blue data points)) as a function of nucleotide position, for a tumor sample.
- a middle plot is of the BAF for two different alleles as a function of nucleotide position, for the tumor sample.
- a bottom plot is the difference between the log Ratio of the two different alleles as a function of nucleotide position and illustrates a partial LOH classification example.
- FIG. 20 illustrates four plots. Two top plots correspond to read coverage as a function of nucleotide position for normal and tumor samples, respectively.
- One bottom plot is of Log Ratio of read coverage in the tumor sample to the read coverage in the normal sample, as a function of nucleotide position.
- the other bottom plot is the difference between log Ratio of the two different alleles as a function of nucleotide position and illustrates a clonal loss classification example.
- a gray line is shown for each of the allele plots and represents the read coverage after a smoothing filter was applied, in these examples, a Savitzky-Golay filter. Smoothing the read coverage allows for less noise in downstream determined coverage features.
- the two-layer clonal LOFI determination process 500 can be used with any number of cancer types to provide decisional support for identifying targeted therapies.
- the FILA-LOFI determination may be performed for patients having a colorectal cancer diagnosis.
- the process 500 may be focused on an FILA-LOFI determination on one specific FILA-A allele (FILA-A * 02:01), and may be used as a companion diagnostic, for example, to a chimeric antigen receptor (CAR) T-cell therapy or another therapy indicated for treatment of tumors having FILA-LOFI (including FILA-LOFI of a specific FILA allele).
- CAR chimeric antigen receptor
- the CAR therapy is targeted to the tumor-specific antigen CEA, a well-known tumor-selective antigen highly expressed in all colorectal cancers and a subset of other epithelial neoplasms.
- the CAR may further comprise a synthetic AND/NOT logic gate system that reacts to two antigens in the body.
- the CAR includes an activating (for example, A module) receptor that can bind to CEA, and a blocking (for example, B module) receptor that blocks T-cell activation and binds to an FILA allele.
- the FILA allele is the FILA-A * 02 allele.
- FILA-A * 02:01 for patients having germline FILA-A * 02:01 expression, if their tumor cells express CEA but have lost expression of FILA-A * 02:01 , the tumor cells will be susceptible to tumor-cell killing by the CAR-T cell described here, but normal cell killing is blocked.
- identifying a sample as having no LOH, clonal LOH, or partial LOH for the HLA-A * 02:01 allele can be used in identifying whether to use CAR therapy or, as may be the case with partial LOH, use a combo therapy that combines a therapy directed at a subpopulation of cancer cells having HLA LOH and another therapy directed at a subpopulation of cancer cells without HLA LOH.
- FIG. 21 illustrates another example process 700 for HLA detection and analysis that may be performed by an HLA and HLA-LOH analysis system, such as that shown in FIG.
- the process 700 which shares features with process 100 in FIG. 2, is able to analyze genetic sequencing data to identify somatic variants in an HLA gene and annotate those variants for their functional effect to provide better information to clinicians for clinical decisions on recommended therapies.
- the process 700 may be used to identify and annotate somatic variants in the HLA Class 1 classical genes, namely HLA-A, HLA-B, and HLA- C in a sample.
- the HLA region is highly polymorphic in nature, hence using a single standard reference genome (hg19 or hg38) may be insufficient to align reads accurately to a patient's HLA loci.
- there is a high level of homology between the 3 classical Class 1 HLA genes and other pseudogenes in the HLA region are highly polymorphic in nature, hence using a single standard reference genome (hg19 or hg38) may be insufficient to align reads accurately to a patient's HLA loci.
- Class 1 HLA type data 702 is obtained.
- Predicted class 1 HLA type data is generated in a process similar to the generation of data 210, as described above. For example, to pull all HLA reads that could potentially map to HLA -A, -B and -C, all patient sequencing reads from the broader HLA loci may be pulled and analyzed by an HLA type predictor (for example, Optitype or Kourami).
- the resulting broader HLA loci data may include reads from non-classical HLA genes (HLA - E, F and G), HLA pseudogenes (N ,J,K and H) and random HLA class 1 gene fragments(L,W,U,T,P,V, etc.).
- An HLA typing module (for example, of a NGS system or other system of patient data) may be used at a process 702 to determine a patient's HLA A, B and C type.
- a patient specific HLA reference file is generated at the process 704, in accordance with examples herein, to provide a reference file to which a patient's HLA reads are aligned.
- Process 704 may comprise processes 124, 126, and/or 128, as described above.
- the index may be built using the Genbank sequences for each HLA allele identified in data 702, as the Genbank sequences include all exons and introns in the HLA Class 1 genes. In alternative embodiments, other similar databases may be used instead of Genbank.
- BAM files for normal tissue sample (706) and tumor tissue sample (708) are analyzed by processes 710 and 712 to determine normal HLA reads and tumor HLA reads, respectively.
- the patient read data from processes 710 and 712 are aligned to the HLA reference file, for example using Novoalign or other sequence alignment process. From the alignment process, aligned reads may then be filtered for variant calling.
- the process 714 may align to a few exons, e.g., exons 2 and 3, whereas in some examples, such as variant calling, the process 714 may align for the entire HLA gene.
- aligned reads may be filtered using a multistep process 716 that processes SAM files to remove multi-mapped reads (for example, reads mapping equally well to more than one HLA allele) and secondary alignment bias.
- the process 716 analyzes the alignment reads from process 714 and identifies and discards duplicate reads, quality control (QC) failed reads (e.g. using QC software such as listed herein), and/or reads where only one of the mates (one of the paired-end reads) is mapped.
- QC quality control
- the process 716 may then count the number of insertion events, deletion events, and mismatch events from a CIGAR string and, from there, compute the total size of insertions, deletions, and mismatches from the CIGAR string.
- event count mismatch event + number of insertions + number of deletions.
- the process 716 passes all reads with an event count greater than 1 , while filtering (discarding) all other reads.
- the process 716 is able to reduce computation time by performing alignment to a reduced reference sequence compared to alignment to all possible sequences, for example, using a reference sequence based on a patient’s determined HLA type or by comparison to normal tissue from non-patients having matched HLA type.
- having processes 714 and 716 and 718 implementable in a modular fashion allows for customizing scripts for variant calling that substantially reduce processing times.
- the process 716 may assign a mapping quality score or value to the filtered reads.
- all filtered reads are assigned the same mapping quality score. This may include, for example, changing the mapping quality to a number that indicates a read that mapped well, having a good quality of mapping to the reference, for example a read that was a good match for the reference sequence.
- the number that indicates good mapping may be approximately 70-100.
- a filtered read s mapping quality value of 20-30 that indicates poor quality of mapping, may be changed to 70.
- a read that aligns equally well to two HLA genes may receive a poor mapping quality score, but for downstream analysis, it may be desirable to have the variant caller treat all filtered reads as equally important, instead of ignoring or giving less weight to filtered reads having a mapping quality value within a particular range. Normalization may be used, because mapping quality may affect downstream variant calling at process 718. In some examples, given that the filtered reads have passed a stringent filtering process as described, a variable mapping quality may have been assigned by Novoalign due to random multi-mapping. Assigning all reads to the same mapping quality removes such variation.
- an optional post-filtering process 717 may be employed to remove non-paired reads.
- the process 717 removes reads if one paired-end read was not mapped, such that both read ends will have to map in order to call a variant in the process 718.
- the process 717 therefore is able to remove mating pairs of reads where, read pairs mate post alignment (e.g., process 714), but post filtering (e.g., process 716), one of the mates has been removed.
- the optional process 717 thus, can be designed to remove the remaining mate.
- the process 717 may remove false mapped reads, which can sometimes occur given the high homology between HLA alleles and HLA genes.
- a variant calling process is performed on the received filtered reads, i.e., the filtered normal and tumor BAM files. For example, using freebayes or other variant calling, the process 718 identifies any nucleotide positions where the patient’s HLA sequences from the filtered normal BAM and filtered tumor BAM files diverge from the HLA reference.
- the process 718 turns off the secondary alignment bit flag in the SAM file, because in the example of Novoalign it only allows one region per read to be assigned as the primary alignment for that read, yet, with the high homology between HLA A, HLA B, and HLA C genes and heterozygous alleles, a single read can be treated as a primary alignment for multiple, different HLA genes.
- the remaining reads are then used for somatic variant calling using Freebayes. Freebayes is not a somatic variant caller, inherently. Therefore, in an example, all possible variants may be called, and then the tumor variant allele frequency (VAF) and the normal VAF may be compared at process 718 to pull the somatic variants.
- VAF tumor variant allele frequency
- a process 720 is provided to apply variant filters to remove low confidence variants, germline variants, and variants likely resulting from sequencing artifacts, for example using techniques described herein.
- the process 720 filters out variants having (1) less than a threshold number of tumor specimen reads that support the variant (in some embodiments, that threshold is 5, in some embodiments that threshold is 4, 3, or 2) or (2) greater than a threshold number of normal specimen sequencing reads that support the variant (in some embodiments, that threshold is 3, in some embodiments that threshold is 2 or 1).
- the process 720 filters out variants having approximately equal VAFs in the normal and tumor specimen.
- the variant if the VAF of the variant in the tumor specimen variant is significantly higher than the VAF of the same variant in the normal specimen, the variant is designated to be a somatic variant. In some embodiments, for one variant, if the VAF in the tumor specimen divided by the VAF in the normal specimen is greater than or equal to a selected VAF ratio threshold value, the variant is designated to be a somatic variant. In some embodiments, the VAF ratio threshold value is 2, in some embodiments the VAF ratio threshold value could be 1.5 or greater. Thus, if tumor VAF divided by normal VAF (the VAR ratio) is greater than threshold value, the variant is designated as somatic. In some examples, a significantly high threshold value is used, which means a VAR ratio greater than 2, 3, 4, or 5, in various applications.
- the VAF can be calculated by dividing the supporting number of reads for a variant by all sequencing reads that map to a location that includes the variant’s location.
- the process 720 could access a panel of normal specimen sequencing data from other patients, which could be a database of germline variants that have been seen in other samples.
- variants that are both listed in the database of germline variants and the patient’s somatic HLA variant list may be removed from the somatic variant list and designated as germline variants.
- the process 720 may filter this normal panel to only include variants associated with the same HLA types as the patient's HLA types (the patient whose normal specimen data is missing).
- multimapping reads (and any reads that align to an HLA gene) may be compared to the reference sequence for that gene to determine if there are any differences (variants), but instead of the variants associated with multimapping reads being more likely to be ignored than the variants associated with other HLA aligned reads, all variants from either category of sequence reads may be analyzed by the process 718, e.g., receiving the same consideration.
- the variant is removed/filtered.
- RPR or all reads map to the left of the variant, indicated by RPL
- this may indicate that the sequencing reads that support the variant are actually reads that are mismapped.
- the process 720 may analyze whether the reads map to a different HLA gene in the HLA reference. If so, the variants may be filtered. For a clinical report, any of these filters may be more stringent. There are various ways to adjust filters to make them more stringent.
- filters at process 720 can be variant filters that call variants having 3 reads at 3% VAF, with approximately 30x coverage, where more stringent filters would be ones that would call variants having 10 reads at 10% VAF, with approximately 100x coverage. If the VAF is approximately 80% or more, the process 720 may identify that to be a germline variant.
- the process 720 applies variant filters that identify variants by forward and reverse strands and variants captured by reads to both the left and the right (RPR and RPL, SAF and SAR).
- the resulting set of variant alleles are provided at the process 722 and may be annotated by a report generator to show their functional effect, including, for example, none (no mutation effect), silent or synonymous mutation, missense mutation, nonsense mutation, loss of function mutation, gain of function mutation, stop codon, frame shift mutation, or other descriptions of types of mutations (variants). Please see Table 2 for an example of the annotation capabilities of the report.
- SnpEff is used to provide the annotation ( hfips://pcin3 ⁇ 4ola.3 ⁇ 4i ⁇ hub.io/Snp£ff/se inputoufpuf/ ).
- Process 722 may further comprise steps described above as processes 16 and 18, which may be used to provide information to a physician to assist with clinical decision-making.
- Table 2 is an example output from the systems and methods. Table 2 describes the type of information that can be associated with each variant detected by one embodiment of the systems and methods.
- Table 3 is another example output from the systems and methods. Table 3 describes the summary information that can be generated for each specimen analyzed by the systems and methods.
- EXAMPLE 3 Rapid detection of somatic variants in human leukocyte antigen class 1 genes from solid tumor samples
- HLA Human Leukocyte Antigens
- Somatic mutations in HLA genes have the potential to reduce the total set of neoantigens available for presentation to T-cells, providing a possible immune escape mechanism for tumors.
- the presence of mutations in HLA genes may thus affect an individual’s response to Immune Checkpoint Blockade (ICB) therapy, as an example.
- IDB Immune Checkpoint Blockade
- the example method takes as input matched tumor and normal sequencing (BAM) files (blocks 708 and 706, respectively) and a patient’s HLA type (block 702).
- the first step involves mapping and aligning reads to the full-length HLA Class 1 genes of an individual patient (block 714). In some examples, this alignment may be performed externally, and merely an output of the alignment is received, e.g., as SAM/BAM files.
- the method filters reads to minimize sequencing errors and further process them to account for potential cross-mapping between genes in the broader HLA locus (block 716).
- the algorithm uses the final set of processed reads from the matched tumor and normal sequencing files to call somatic variants (block 718). The final set of variants are then provided with a genetic variant annotation that summarizes their functional effect (block 722).
- the sequencing data were accessed from a public database (for example, TCGA). In other examples, the sequencing data was accessed from a proprietary or private database.
- Validation may comprise one or more of the following methods:
- validation may include using the systems and methods to analyze DNA sequencing data associated with a tumor specimen and a normal specimen from the same subject (for example, patient) to detect variants in HLA genes in the tumor specimen, then analyzing RNA sequencing data (for example, a FASTQ file or files) associated with the tumor specimen to determine whether the RNA sequencing data has any sequencing reads that support the variant.
- RNA sequencing data for example, a FASTQ file or files
- 70 variants that were detected in DNA data 54 of those variants were also detected (were covered by at least one supporting sequencing read) in RNA data by the systems and methods.
- 70 variants that were not detected in DNA data were tested and were not detected (were covered by 0 supporting sequencing reads) in the RNA data.
- validation may include using the systems and methods to detect variants in HLA genes in sequencing data (tumor and normal DNA sequencing data) associated with a specimen to generate a first variant report or report portion, using a second method of HLA variant detection to process the same sequencing data to generate a second variant report or report portion, and comparing the two variant reports or report portions to determine their degree of similarity.
- the second method of HLA variant detection is Polysolver and the sequencing data are TCGA data. Of 46 variants detected by Polysolver, the systems and methods detected 41. For 3 of the 5 variants that were detected by Polysolver but not detected by the systems and methods, the quality of the supporting sequencing reads were below the threshold set by Freebayes.
- FIG. 22 is a chart comparing the VAF associated with the variant detected by the systems and methods (x-axis) to the VAF associated with the variant detected by Polysolver, and the results were highly correlated, showing a Spearman correlation coefficient of approximately 0.91.
- validation may include using the systems and methods to detect variants in HLA genes in short read sequencing data associated with a specimen to generate a first variant report, using a second method of variant detection to process long-read sequencing data from the same specimen to generate a second variant report, and comparing the two variant reports to determine the degree of similarity of the reports.
- validation may include detecting non-sense mediated decay resulting from nonsense mutations (NM).
- NMD Nonsense Mediated Decay
- PTC Premature Termination Codons
- FIG. 23 is a bar graph comparing expression levels (y-axis) between an allele with PTC and an allele without PTC (see legend) for the same HLA gene in a TCGA sample (sample indicated along the x-axis).
- the plot compared the allele with PTC to the allele without PTC in a heterozygous locus for 6 samples, and 4 out of 6 samples showed strong evidence of NMD. Strong evidence of NMD is defined as the allele with NMD having less than half the expression of the allele without NMD in a heterozygous locus
- FIG. 24 is a bar graph comparing expression levels (y-axis) between different alleles for an example sample - TCGA-CV-6937.
- This sample has a nonsense mutation in the A2 allele (a premature stop codon/PTC). It is observed that A2’s expression (orange bar) is only 10% of A1 ’s (blue bar), which can be explained by nonsense mediated decay.
- the HLA-B and C alleles do not have a PTC mutation and they do not exhibit effects of NMD, as expected.
- FIG. 25 is a plot of expression data showing RNA allelic imbalance as a function of HLA status in an example implementation of the present techniques.
- FIG. 25 shows the allele-level HLA expression from TCGA data, with nonsense (NM) mutations, without nonsense mutations (No NM), with loss of heterozygosity (LOH), and with no LOH.
- HLA genes with nonsense mutations had a similar expression profile to genes with LOH, a well-known immune evasion mechanism.
- the present techniques can be applied in examples in which a normal patient sample is not input. Instead, in some examples, a tumor-only analysis may be performed where data from the normal sample(s) is replaced with historical data from a large database of previously run samples with the same HLA genotype. [284] If the patient’s normal sample has allelic imbalance, it may be beneficial for process 118 to compare the cancer sample from a patient to one or more normal samples collected from people who are not the patient (for example, a “panel of normal samples”).
- process 118 may use a large historical dataset (for example, a proprietary or public dataset) to compare the tumor sample to the median of all normal samples with the same haplotype.
- a large historical dataset for example, a proprietary or public dataset
- a panel of normals may have the median coverage of A * 02:01 and the other allele at each position for each allele when paired with A * 02:01 :
- the panel of normals includes 6,005 HLA-A * 02:01 heterozygous samples.
- the frequency of the allele paired with A * 02:01 (only alleles with more than 20 samples were kept) is shown in Table 5:
- FIG. 26 illustrates plots of different metrics for a panel of normal samples (left) and a patient tumor sample (right) for a stable A * 02:01/A * 03:01 sample.
- Wide lines represent the coverage, B-allele fraction or log ratio coverage in the normal (left) or tumor (right) sample.
- the thin lines represent the coverage of the panel of normals and B-allele fraction or log ratio coverage calculated using the panel of normals.
- the HLA LOH module relies very heavily on comparing the coverage of each allele in the normal and the tumor sample to determine if an allele is lost and may result in false positives.
- the techniques herein are more accurate and precise when run in tumor-only mode, facilitated by a database including at least 90,000 previous HLA LOH calls to use when replacing the normal sample(s) with a database of previously run normals.
- B-allele fraction ratio the ratio of the B-allele fraction in the tumor to the B-allele fraction in the normal sample
- Fig. 27 illustrates a plot of B-allele fraction of the normal and tumor sample in -6,000 HLA-A * 02:01 heterozygous samples. Colors indicate the LOH call from an example implementation of a variant calling process described herein.
- samples with a B-allele fraction close to 50% in the tumor sample can be classified as partial or clonal losses if the B-allele fraction of the normal sample is low/high enough.
- This condition can occur in a non-negligible number of samples as the distribution of the median B-allele fraction in normal samples has a relatively large standard deviation, with tails under 40% and over 60%.
- FIG. 28 illustrates a distribution of B-allele fraction in normal samples.
- FIG. 29 illustrates a frequency of partial loss calls by B-allele fraction percentile (each bin contains 600 samples).
- FIG. 30 illustrates a frequency of clonal loss calls by B-allele fraction percentile (each bin contains 600 samples)
- Allelic imbalance is haplotype and position specific.
- FIG. 31 shows the median B-allele fraction of HLA-A * 02:01 in the normal sample when paired with other frequent alleles (> 30 samples).
- FIGS. 32A and 32B illustrate the BAF of HLA-A * 02:01 by position in the HLA-A * 02:01 / A:03:01 haplotype, FIG. 32A normal samples and FIG. 32B tumor samples.
- FIG. 33 illustrates a median BAF and 95% confidence interval of BAF of HLA-A * 02:01 by position in the HLA-A * 02:01 / A:03:01 haplotype [296]
- the allelic imbalance in the normal sample is not only a general bias but we also see, in some examples, different artifacts depending on the position on the FILA gene.
- FIG. 33 illustrates the B-allele fraction of FILA-A * 02:01 in the normal sample by the position, when paired with FILA-03:01 (on 1 ,057 samples). Green zones are exon 2 and exon 3.
- FIGS. 34A and 34B illustrate FILA-A LOFI status in function of the BAF of A * 02:01 in the normal and tumor sample in the 6,005 A * 02:01 heterozygous samples when using the patient’s normal (FIG. 34A) or the panel of normals (FIG. 34B).
- FIGS. 35A-35D illustrate a frequency of partial loss (FIGS. 34a and 34B) or clonal loss (FIGS. 34C and 34D) calls by B-allele fraction percentile (each bin contains 600 samples) in the research use only (RUO) version of the algorithm (FIGS. 34A and 34C) or the CDx version (FIG. 34B and 34D).
- B-allele fraction percentile each bin contains 600 samples
- RUO research use only
- FIGS. 34A and 34C the CDx version
- Table 6 provides a detailed breakdown of calls:
- routines, subroutines, applications, or instructions may constitute either software (e.g., code embodied on a machine-readable medium or in a transmission signal) or hardware.
- routines, etc. are tangible units capable of performing certain operations and may be configured or arranged in a certain manner.
- one or more computer systems e.g., a standalone, client or server computer system
- one or more hardware modules of a computer system e.g., a processor or a group of processors
- software e.g., an application or application portion
- a hardware module may be implemented mechanically or electronically.
- a hardware module may comprise dedicated circuitry or logic that is permanently configured (e.g., as a special-purpose processor, such as a microcontroller, field programmable gate array (FPGA) or an application-specific integrated circuit (ASIC)) to perform certain operations.
- a hardware module may also comprise programmable logic or circuitry (e.g., as encompassed within a processor or other programmable processor) that is temporarily configured by software to perform certain operations. It will be appreciated that the decision to implement a hardware module mechanically, in dedicated and permanently configured circuitry, or in temporarily configured circuitry (e.g., configured by software) may be driven by cost and time considerations.
- the term "hardware module” should be understood to encompass a tangible entity, be that an entity that is physically constructed, permanently configured (e.g., hardwired), or temporarily configured (e.g., programmed) to operate in a certain manner or to perform certain operations described herein.
- hardware modules are temporarily configured (e.g., programmed)
- each of the hardware modules need not be configured or instantiated at any one instance in time.
- the hardware modules comprise a processor configured using software
- the processor may be configured as respective different hardware modules at different times.
- Software may accordingly configure a processor, for example, to constitute a particular hardware module at one instance of time and to constitute a different hardware module at a different instance of time.
- Flardware modules can provide information to, and receive information from, other hardware modules. Accordingly, the described hardware modules may be regarded as being communicatively coupled. Where multiple of such hardware modules exist contemporaneously, communications may be achieved through signal transmission (e.g., over appropriate circuits and buses) that connects the hardware modules. In embodiments in which multiple hardware modules are configured or instantiated at different times, communications between such hardware modules may be achieved, for example, through the storage and retrieval of information in memory structures to which the multiple hardware modules have access. For example, one hardware module may perform an operation and store the output of that operation in a memory device to which it is communicatively coupled. A further hardware module may then, at a later time, access the memory device to retrieve and process the stored output.
- Flardware modules may also initiate communications with input or output devices, and can operate on a resource (e.g., a collection of information).
- a resource e.g., a collection of information.
- the various operations of the example methods described herein can be performed, at least partially, by one or more processors that are temporarily configured (e.g., by software) or permanently configured to perform the relevant operations. Whether temporarily or permanently configured, such processors may constitute processor-implemented modules that operate to perform one or more operations or functions.
- the modules referred to herein may, in some example embodiments, comprise processor-implemented modules.
- the methods or routines described herein may be at least partially processor-implemented. For example, at least some of the operations of a method can be performed by one or more processors or processor-implemented hardware modules. The performance of certain of the operations may be distributed among the one or more processors, not only residing within a single machine, but also deployed across a number of machines. In some example embodiments, the processor or processors may be located in a single location (e.g., within a home environment, an office environment or as a server farm), while in other embodiments the processors may be distributed across a number of locations.
- the performance of certain of the operations may be distributed among the one or more processors, not only residing within a single machine, but also deployed across a number of machines.
- the one or more processors or processor- implemented modules may be located in a single geographic location (e.g., within a home environment, an office environment, or a server farm). In other example embodiments, the one or more processors or processor-implemented modules may be distributed across a number of geographic locations.
- any reference to “one embodiment” or “an embodiment” means that a particular element, feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment.
- the appearances of the phrase “in one embodiment” in various places in the specification are not necessarily all referring to the same embodiment.
- Some embodiments may be described using the expression “coupled” and “connected” along with their derivatives. For example, some embodiments may be described using the term “coupled” to indicate that two or more elements are in direct physical or electrical contact. The term “coupled,” however, may also mean that two or more elements are not in direct contact with each other, but yet still co-operate or interact with each other. The embodiments are not limited in this context.
- the terms “comprises,” “comprising,” “includes,” “including,” “has,” “having” or any other variation thereof, are intended to cover a non-exclusive inclusion.
- a process, method, article, or apparatus that comprises a list of elements is not necessarily limited to only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
- "or” refers to an inclusive or and not to an exclusive or. For example, a condition A or B is satisfied by any one of the following: A is true (or present) and B is false (or not present), A is false (or not present) and B is true (or present), and both A and B are true (or present).
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- General Health & Medical Sciences (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Pathology (AREA)
- Molecular Biology (AREA)
- Analytical Chemistry (AREA)
- Biophysics (AREA)
- Genetics & Genomics (AREA)
- Chemical & Material Sciences (AREA)
- Proteomics, Peptides & Aminoacids (AREA)
- Bioinformatics & Computational Biology (AREA)
- Biotechnology (AREA)
- Evolutionary Biology (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Theoretical Computer Science (AREA)
- Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US202163203753P | 2021-07-29 | 2021-07-29 | |
PCT/US2022/038948 WO2023009863A1 (en) | 2021-07-29 | 2022-07-29 | Detection of genetic variants in human leukocyte antigen genes |
Publications (1)
Publication Number | Publication Date |
---|---|
EP4377479A1 true EP4377479A1 (de) | 2024-06-05 |
Family
ID=85088032
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
EP22850393.4A Pending EP4377479A1 (de) | 2021-07-29 | 2022-07-29 | Nachweis genetischer varianten in menschlichen leukozytenantigengenen |
Country Status (2)
Country | Link |
---|---|
EP (1) | EP4377479A1 (de) |
WO (1) | WO2023009863A1 (de) |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140065620A1 (en) * | 2011-12-29 | 2014-03-06 | Mayo Foundation For Medical Education And Research | Nucleic acids for detecting breast cancer |
EP3765636A1 (de) * | 2018-03-14 | 2021-01-20 | Université de Versailles Saint-Quentin-en-Yvelines | Verfahren zur identifizierung einer oder mehrerer mutationen in einer hotspot-mutationssequenz |
AU2020221847A1 (en) * | 2019-02-12 | 2021-09-16 | Tempus Ai, Inc. | Detection of Human Leukocyte Antigen loss of heterozygosity |
-
2022
- 2022-07-29 WO PCT/US2022/038948 patent/WO2023009863A1/en active Application Filing
- 2022-07-29 EP EP22850393.4A patent/EP4377479A1/de active Pending
Also Published As
Publication number | Publication date |
---|---|
WO2023009863A1 (en) | 2023-02-02 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11081210B2 (en) | Detection of human leukocyte antigen loss of heterozygosity | |
US10975445B2 (en) | Integrated machine-learning framework to estimate homologous recombination deficiency | |
EP4073805B1 (de) | Systeme und verfahren zum vorhersagen des mangels an homologer rekombination bei einer probe | |
US11475978B2 (en) | Detection of human leukocyte antigen loss of heterozygosity | |
US20200210852A1 (en) | Transcriptome deconvolution of metastatic tissue samples | |
US20220215900A1 (en) | Systems and methods for joint low-coverage whole genome sequencing and whole exome sequencing inference of copy number variation for clinical diagnostics | |
WO2017065959A2 (en) | Methods and compositions that utilize transcriptome sequencing data in machine learning-based classification | |
WO2022109607A2 (en) | Determination of cytotoxic gene signature and associated systems and methods for response prediction and treatment | |
US20210398617A1 (en) | Molecular response and progression detection from circulating cell free dna | |
US20230064530A1 (en) | Detection of Genetic Variants in Human Leukocyte Antigen Genes | |
WO2023277932A1 (en) | Detection of human leukocyte antigen loss of heterozygosity | |
EP4377479A1 (de) | Nachweis genetischer varianten in menschlichen leukozytenantigengenen | |
Feldhahn | Computational methods for personalized cancer therapy based on genomics data |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE INTERNATIONAL PUBLICATION HAS BEEN MADE |
|
PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE |
|
17P | Request for examination filed |
Effective date: 20231214 |
|
AK | Designated contracting states |
Kind code of ref document: A1 Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR |