HLA typing methods

Abstract

The present invention relates to a method for characterizing the HLA status of a genetic sample obtained from a subject, comprising the steps of: i. performing DNA or RNA sequencing on the genetic sample obtained from said subject; ii. aligning the obtained sequences to one or more reference HLA allele sequences; and iii. determining the HLA status by applying variant calling techniques to identify the presence or type of variants in the HLA sequences of the genetic sample.

Description

[Technical field]

[0001] FIELD OF THEINVENTION The present invention relates to HLA typing methods for applications in HLA matching in organ transplantation, hematopoietic stem cell transfusion, disease association testing, prophylactic and therapeutic vaccines in cancer and infectious diseases, and finally predicting response to immunotherapy, especially in cancer. [Background technology]

[0002] 2. Background of the Invention Full-length human leukocyte antigen (HLA) typing remains a continuing challenge, as it represents one of the most complex polymorphic regions in the human genome. 1~3 Classical HLA proteins bind in complexes with peptides that can be presented on the cell surface. Once presented on the cell surface, these complexes can then be recognized by effector T cells of the adaptive immune system. Since class I HLA proteins present peptides on the surface of all human cells, precise HLA genotyping is of critical importance in organ transplantation, especially in organ transplantation where donor and recipient HLA matching is required. 4~6 Accurate knowledge of an individual's HLA genotype is also important in disease association studies, where HLA allelic variants have strong genetic associations with many common human diseases. 7 In addition, variations in HLA alleles may affect drug sensitivity. 9 and susceptibility to adverse drug responses 10 In addition, many studies have frequently linked it to disease susceptibility. 4、8 In cancer, certain HLA genotypes 10 In particular, it has been demonstrated that the diversity of a patient's HLA genotype can predict response to immune checkpoint inhibitors (ICIs). 11~13 .

[0003] The clinical importance of accurate HLA typing is very well established, and next-generation sequencing (NGS) data has recently been adopted by many diagnostic laboratories as the preferred source of data for performing reliable HLA typing. 14、15 The primary outcome of HLA typing is the assignment of unique HLA names (e.g., HLA-A*02:01:01:01), termed HLA alleles, constituting a resolution of up to four fields separated by colons. The four fields of this HLA nomenclature represent (1) allelic groups, (2) protein groups, (3) synonymous DNA changes within protein-coding regions, and (4) variants in noncoding regions.

[0004] NGS-based HLA typing methods can currently be divided into two categories: HLA-targeted sequencing with high sequencing depth (e.g., PCR-based target amplification) and standard NGS with medium sequencing depth (e.g., whole exome sequencing (WES), whole genome sequencing (WGS), and RNA sequencing (RNA-Seq)). HLA-targeted sequencing (or deep sequencing) of HLA regions is the most information-dense type of NGS and is therefore most often used to resolve HLA typing ambiguities of exons encoding peptide-binding clefts (and is the most informative in determining the first and second fields of resolution). This is due to the critical importance of determining donor HLA-peptide complex presentation so that tissue compatibility can be achieved in transplantation. Targeted HLA sequencing is performed on a small portion of the sample. 17 PCR amplification 16 , which is considered the gold standard HLA typing for clinical use, although this is arguably 14 However, compared to standard NGS and especially WES-based approaches, it is more labor-intensive, expensive, and time-consuming. 14Therefore, targeted HLA sequencing is mostly used to elucidate HLA genotypes and ambiguities of peptide-binding cleft exons. Therefore, most of the described HLA alleles have incomplete sequences with enhanced coverage of the binding cleft exons, and only a small number of alleles are accompanied by complete full-length HLA sequences. 18、19 .

[0005] This lack of full-length HLA sequence typing is suboptimal since identification of the complete HLA region at a resolution of four fields has important medical and genomic applications. Full-length HLA sequence typing is useful, for example, to perform ancestry-based analyses. 20 , HLA-based disease association testing has been shown to be critically important in identifying causative variants 21 The importance of full-length HLA sequence typing has also been demonstrated to help optimize donor selection, improve clinical outcomes, and reduce transplant complications, as clearly demonstrated in hematopoietic cell transplantation (HCT). 22~26 Furthermore, full-length HLA typing may provide novel insights into the regulation of HLA gene transcript expression, including epigenetic mechanisms, leading to improved understanding of combined immune disorders. 27、28 .

[0006] The advent of widely available standard NGS data has led to an increase in the number of computational NGS-based HLA typing solutions. 16、19、29~32 However, the majority of these tools perform HLA typing by identifying the closest matching HLA alleles via sequence alignment of NGS reads against reference sequences in the IPD-IMGT / HLA database. 1 Unfortunately, the IPD-IMGT HLA database is sparse in its coverage of complete alleles typed in the third and fourth fields of resolution, so standard NGS-based HLA typing solutions do not reliably provide complete HLA genotypes or allow the discovery of novel HLA alleles. Importantly, some of these computational methods19 does provide the ability to output full-length HLA genotypes using WES data, but until very recently 33~35 However, reliable benchmarking data was only available for protein coding sequences. De novo assembly methods are not limited to relying on the IPD-IMGT HLA database and may therefore have the potential to identify novel HLA alleles. 32、36 However, these tools are computationally expensive and their accuracy depends on the need to use long reads for deep coverage of the HLA region and proper phasing.

[0007] Therefore, there is a need for improved methods and techniques to perform efficient and accurate HLA typing to characterize an individual's HLA status for therapeutic assessment and intervention. Summary of the Invention [Problem to be solved by the invention]

[0008] Summary of the Invention The present invention provides an HLA typing solution based on alignment to known HLA sequences (e.g., from the IPD-IMGT HLA database), but at the same time allows for the discovery of novel HLAs by utilizing germline and / or somatic variant calling methods. This personalized HLA typing method is applied to sequencing data and demonstrates the ability to identify novel HLA alleles and resolve HLA ambiguity, especially at the third and fourth fields of resolution. [Means for solving the problem]

[0009] According to a first aspect of the present invention there is provided a method for characterising the HLA status of a genetic sample obtained from a subject, comprising: i. performing DNA or RNA sequencing on the genetic sample obtained from the subject; ii. aligning the obtained sequences to one or more reference HLA allele sequences; iii. determining HLA status by applying variant calling techniques to identify the presence or type of variants in the HLA sequences of said genetic sample; There is a method including the steps:

[0010] Description of the drawings The present invention will now be described with reference to the accompanying drawings. [Brief description of the drawings]

[0011] [Figure 1] Illustrating the steps taken in carrying out the present invention, NGS reads are used to perform patient-tailored HLA typing in an HLA-centric germline calling approach. [Diagram 2] Exemplifying an in silico allele experiment, after typing the closest known HLA alleles, an individual's HLA-A is spiked-in using Bam Surgeon with a comprehensive set of different variant types, including single nucleotide variations, insertions, and deletions. [Figure 3a] 1 shows visualization of germline variants matching the detected novel alleles, (a) represents a single nucleotide polymorphism (SNP) affecting the 3'UTR region of the HLA*B allele. [Figure 3b] A visualization of germline variants matching the detected novel alleles is shown, (b) represents an insertion affecting the first intron of the HLA*DRB1 allele. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Detailed Description The present invention discloses an HLA typing approach that relies on alignment of a reference library (e.g., IPD-IMGT HLA database) to known HLA alleles, but at the same time allows the discovery of new HLA alleles. It therefore provides a database matching-based method that is also capable of discovering new HLA alleles. The vast majority of HLA typing tools that use the IPD-IMGT HLA database are limited to typing only known HLA alleles. Therefore, the completeness of HLA typing by database matching-based methods relies on the completeness of the IPD-IMGT HLA database. This limitation is particularly problematic for HLA typing at the third and fourth fields of resolution.

[0013] The method of the present invention characterizes the HLA status of a genetic sample obtained from a subject, comprising: i. performing DNA or RNA sequencing on the genetic sample obtained from the subject; ii. aligning the obtained sequences to one or more reference HLA allele sequences; iii. determining HLA status by applying variant calling techniques to identify the presence or type of variants in the HLA sequences of said genetic sample; by.

[0014] The term "DNA or RNA sequencing" refers to any method of sequencing DNA or RNA. DNA sequencing provides a genetic profile of a sample, while RNA sequencing reflects only the sequences that are actively expressed in the sample. DNA or RNA sequencing techniques include whole genome sequencing, exonic sequencing, Sanger sequencing, and targeted HLA sequencing.

[0015] The term "exonic sequencing" refers to the sequencing of the protein-coding regions of genes in the genome. It is often referred to as "exome sequencing" or "whole exome sequencing (WES)", and such terms may be used interchangeably herein.

[0016] Techniques for performing such sequencing are well known in the art. Typically, such techniques rely on the selection of only a subset of protein-coding DNA, called exons, and subsequent sequencing of the exonic DNA with high-throughput DNA sequencing tools.

[0017] Many different high-throughput sequencing tools exist for performing exonic sequencing, and the present invention is not limited by the technique used.

[0018] There are many next-generation sequencing platforms available from various manufacturers, each capable of performing massively parallel exome sequencing, including the Roche 454 sequencer, the Life Sciences SOLID sequencer, the Life Technologies Ion Torrent sequencer, and the Illumina MiSeq, HiSeq, and NovaSeq platforms.

[0019] As an alternative to exonic genome sequencing, "whole genome sequencing" may be used, which involves sequencing the entire genome of a subject and may be performed using the same platforms as above.

[0020] Another alternative to exonic genome sequencing is "Sanger sequencing." The terms "Sanger sequencing" and "chain termination sequencing" may be used interchangeably and refer to the sequencing of DNA. Techniques for performing such sequencing are well known in the art.

[0021] The techniques typically involve selective incorporation of chain-terminating dideoxynucleotides by DNA polymerase during DNA replication, resulting in the generation of chain-terminating oligonucleotides due to the cessation of DNA replication. Subsequent separation of the chain-terminating oligonucleotides, such as by gel electrophoresis, allows the determination of the DNA sequence.

[0022] Another alternative to exonic genome sequencing is "targeted HLA sequencing." The term "targeted HLA sequencing" refers to the sequencing of HLA regions of genes in the genome.

[0023] However, in the present invention, preference is given to performing exonic sequencing.

[0024] The term "determining HLA status" refers to either the identification of germline variants in a genetic sample obtained from a subject (but a subject may also be a patient) or the identification of somatic variants in a genetic sample obtained from a tumor patient.

[0025] After sequencing, the reads from the sequencing are aligned to a reference HLA allele sequence. The alignment is performed using one or more reference HLA allele sequences.

[0026] In the context of germline variant calling, raw sequence data are aligned to a standard reference sequence for the species of interest. Alignment to the reference sequence allows for the identification of genotypes based on the assumption that most genomes are diploid.

[0027] In one embodiment, the alignment can be performed against HLA sequences contained in an accessible HLA reference library. In a preferred embodiment, the alignment is performed against sequences contained in the database of the International Immunogenetics Information System (IMGT; www.imgt.org).

[0028] The IPD-IMGT HLA database of known HLA sequences is described in Robinson et al., Nuc Acids Res. 43, D423-31 (2015). The library can be accessed at www.ebi.ac.uk / ipd / imgt / hla, along with the experiments disclosed herein using release v2.0.10.

[0029] Techniques and software for performing DNA or RNA sequence alignment with a reference sequence are known in the bioinformatics arts. Exemplary software programs capable of performing the alignment will be familiar to those of skill in the art. Such software includes BWA (arXiv:1303.3997[q-bio.GN]), GSNAP (Thomas et al., Bioinformatics 2010 26: 873-881), and Maq (Heng et al., Genome Research 2008 18: 1851-1858). Other suitable software packages will be familiar to those of skill in the art.

[0030] In a preferred embodiment, the alignment is performed using GSNAP.

[0031] In a preferred embodiment, alignment with reference HLA alleles is performed using a method known as OncoHLA (Sverchkova et al. HLA 94, 504-513, 2019-ref. 18), available from NEC-OncoImmunity. The technique utilizes alignment of reads to a library of reference alleles, followed by allele type determination based on the distribution of aligned reads and prior probability of ethnic frequency of the allele. It incorporates GSNAP for alignment as well as providing an estimate of allele type based on the prior probability of aligned reads and ethnic frequency of the allele.

[0032] In the context of somatic variant calling, the human reference genome characterizes each HLA gene with a single sequence, which may misrepresent the true HLA alleles of individual subjects. Therefore, to accurately detect somatic mutations in HLA genes from tumor-associated tissues, the reconstructed germline HLA sequence of the subject's own sample may be used as a reference sequence for alignment. For example, germline variant calling may be performed on normal matched tissues from patients, such as normal peripheral blood mononuclear cells (PBMCs), to provide a reference for somatic variant calling on tumor-associated tissues.

[0033] In one embodiment, the reference germline HLA sequences are preferably reconstructed using the methods detailed in the present claimed invention.

[0034] In another embodiment, the reference germline HLA sequences may be reconstructed using a modified version of the method detailed in the present claimed invention.

[0035] After alignment, the aligned reads are input into a variant calling tool.

[0036] The term "germline variant calling" refers to the detection of sequence variants in a genetic sample derived from a patient (eg, from PBMCs).

[0037] The term "somatic variant calling" refers to the detection of sequence variants within the genetic information of somatic cells.

[0038] Specifically, variant calling detects where the reads differ from the genome in the target region. This can be in conjunction with single nucleotide polymorphisms (SNPs) and short insertions and deletions (indels), as will be appreciated by those skilled in the art. In one embodiment, the methods of the invention utilize one or more separate germline variant calling software tools to detect only those variants that are selected for analysis.

[0039] In a preferred embodiment, germline variant calling is performed to select variants with a read depth of >10 and a variant allele frequency of >0.30. Software tools are publicly available for performing germline variant calling, including the Genome Analysis Toolkit GATK-Haplotypecaller (DePristo et al., Nat Genet. 2011; 43(5): 491-8), Platypus (Rimmer et al., Nat Genet. 2014; 46(8):912-8), DRAGEN (Illumina), DeepVariant (Verily Life Sciences), and Strelka2 (Kim et al., Nat Methods 2018 15: 591-594). Each can be performed according to the developer's protocol.

[0040] GATK is the preferred method for performing germline variant calling because it identifies differences between reads and reference sequences using state-of-the-art statistical methods (e.g., logistic regression, hidden Markov models, and naive Bayes classification) to detect sequence differences.

[0041] Platypus is a haplotype-based method for germline variant calling that achieves high sensitivity and specificity using local realignment and local assembly of reads for efficient and accurate variant detection.

[0042] The DRAGEN platform utilizes highly reconfigurable field programmable gate arrays to provide hardware-accelerated implementations of genomic analysis algorithms, realized as logic circuits, that provide near-instantaneous output to detect small variants with high analytical sensitivity and specificity.

[0043] DeepVariant is a deep learning-based method for germline variant calling supported in diploid organisms. The method generates pile-up image tensors of aligned reads and then classifies each tensor before reporting the reads.

[0044] Strelka2 is a variant calling method that introduces a novel mixture model-based estimation of indel error parameters from each sample, an efficient stepwise haplotype modeling strategy, and a normal sample contamination model to improve liquid tumor analysis.

[0045] In one embodiment, the methods of the invention are performed using six separate somatic variant calling software tools, with only high quality candidates detected by at least three of the six software tools selected for analysis.

[0046] In a preferred embodiment, somatic variant calling is performed to select variants with a read depth of >10 and a minimum number of alternative (mutant) reads of 3 in the tumor data and 0 in the normal data.

[0047] Software tools are publicly available for performing somatic variant calling, including Platypus (Rimmer et al., Nat Genet. 2014; 46(8):912-8) and Strelka2 (Kim et al., Nat Methods 2018 15: 591-594).

[0048] In addition to germline and somatic variant calling, the method of the present invention may incorporate a subsequent step of haplotype estimation (phasing), which is performed to identify HLA variants. A haplotype is two sequences of alleles inherited together from an individual's parents, and it is desirable to split the genotype into two haplotypes.

[0049] Statistical methods used for haplotype inference are known, and computational approaches are available, including WhatsHap. 36 .

[0050] The haplotype resolution can then be used to determine protein haplotypes using other bioinformatics tools. For example, 37 can be used to calculate protein haplotypes from phase variation data. Other suitable bioinformatics tools will be apparent to those of skill in the art.

[0051] In one embodiment, customization features may be developed to extend the capabilities of other bioinformatics tools to annotate and fully reconstruct candidate variant HLA gene, transcript, and protein sequences.

[0052] The method can be performed to determine variations in protein coding regions, but can also be performed to determine variations in non-coding regions.

[0053] The sample obtained from the subject used in the method of the present invention can be any suitable biological sample. In a preferred embodiment, the sample is obtained from tissue for transplantation (donor sample) or from blood. In some embodiments, the sample is obtained from tumor-associated tissue, e.g., a cancer biopsy.

[0054] The present invention is particularly useful in organ transplantation, where adaptive immunity is the primary response exerted against the transplanted tissue. HLA typing is therefore a key requirement for reducing rejection rates or minimizing side effects. This is particularly important in bone marrow and kidney transplantation. The selection of optimal donors is influenced by HLA typing, which has a positive effect on graft acceptance if the donor and recipient are matched with respect to MHC antigens.

[0055] According to the present invention there is a method for determining suitability for organ transplantation, comprising determining the HLA status of both the recipient and the donor by carrying out the method described above or according to claim 1 and its dependent claims, and comparing the HLA status of the recipient and the donor to identify a suitable HLA match.

[0056] The method is preferably carried out to identify matches for MHC class I and MHC class II, in particular HLA-A, HLA-B, HLA-C, and HLA-DR.

[0057] The method of the invention may also be used to determine whether a subject will respond to cancer immunotherapy, in which the HLA status of the subject is determined by carrying out a method disclosed above or according to claim 1 or any of its dependent claims and comparing the HLA status with specific HLA genotypes known to respond to immunotherapy.

[0058] Chowell et al., Science, 2018; 359(6375):582-587 disclose that the greater the diversity (e.g., heterozygosity) within a particular HLA molecule (e.g., class-I or HLA-I), the greater the repertoire of tumor-derived neoantigens that can be presented to cytotoxic T cells. Conversely, homozygosity at at least one HLA class I locus was associated with reduced survival to checkpoint inhibitor therapy, as individual subjects homozygous for at least one HLA-I locus would be predicted to present a smaller, less diverse repertoire of tumor-derived neoantigens to cytotoxic T cells. Somatic mutations in HLA have also been shown to be indicative of patient response.

[0059] Thus, the methods of the invention can be used to assess whether a patient undergoing or intended to undergo immunotherapy, such as checkpoint inhibitor therapy (e.g., with a PD-1 / PD-L1 or CTLA4 targeted therapeutic agent), will respond to the therapy. Identifying patients who are homozygous for at least one HLA-I locus will be predictive of reduced response to such therapy. EXAMPLES

[0060] Working Example HLA database closest match typing from WES data An overview of the workflow of the personalized HLA typing and tumor-specific HLA variant calling pipeline is illustrated in Figure 1. Briefly, the enormous complexity of the HLA region leads to inaccurate HLA typing with traditional mapping approaches to the reference genome. This complexity is resolved either by aligning NGS reads to an HLA reference sequence library or by applying de-novo assembly methods. The method outlined in Figure 1 is based on the former and uses the IPD-IMGT HLA database of known HLA sequences. 2 HLA typing is performed by aligning the reads to the 19 Once the WES reads were aligned to all known HLA alleles available in IMGT, the HLA alleles were determined by the OncoHLA HLA typing tool using an integer linear programming (ILP) algorithm that uses prior probabilities of allelic ethnic frequencies. The output of OncoHLA includes the closest matching HLA alleles from the HLA database and the associated HLA sequence at up to four fields of resolution for each allele.

[0061] Integration of germline variant calling to achieve personalized HLA typing from WES data Reads from the WES data were aligned to the reference HLA sequences output by OncoHLA (see Figure 1). 38(Version 2020-05-30) has high accuracy in mapping NGS reads to highly polymorphic regions. 39 The resulting alignment files were processed according to standard practices (including read sorting and duplicate marking) and analyzed using two state-of-the-art germline variant calling tools, GATK-HaplotypeCaller and GATK-HaplotypeCaller. 40 (v4.0.6) and Strelka2 41 (v2.0.10). In this step, each candidate germline variant was determined to be associated with a potential mistyped or novel HLA allele. To reduce false positive calls, only variants detected by both variant callers and with a read depth >10 and variant allele frequency >0.30 were considered for further analysis.

[0062] Integration of ensemble somatic variant calling for identification of tumor-specific HLA variants Six different state-of-the-art somatic variant calling tools were used to identify somatic mutations in the HLA region by aligning matched tumor-normal WES reads in the HLA region against previously typed personalized HLA sequences (known or potentially novel) (see Figure 1). Due to the high complexity of detecting somatic variants in HLA alleles, only high-quality candidate variants reported by at least three of the six variant calling tools were considered. 41~46 Each tool has its own algorithm and unique rule set to distinguish variants from background noise, so they are combined to reduce the false positive detection rate. Several additional filters were also applied to discard potential false positive calls, including a minimum read depth of 10 in both tumor and normal data at the variant location and a minimum number of alternative (mutant) reads of 3 in tumor and 0 in normal data. Ensembl Variant Effect Predictor (VEP) was then used to assess the impact of the detected variants on the resulting gene product. 47The toolkit was used.

[0063] Individualized HLA alleles with phase-appropriate haplotypes and reconstruction of tumor-specific HLA alleles Identification of the correct genomic phase (or haplotype) of the HLA variants was performed for subsequent accurate reconstruction of well-phased candidate variant HLA sequences. To determine the phase relationship between heterozygous variants along two target HLA alleles, WhatsHap 36 (v0.17) was used. Once phasing was performed, Ensembl Variant Effect Predictor (VEP) was then used to assess the functional impact of the detected variants in the HLA allele sequences. 47 (v95) How to embed Haplosaurus 37 was used. Customization features were developed to extend Haplosaurus capabilities to annotate and fully reconstruct candidate variant HLA gene, transcript, and protein sequences. Once variant HLA sequences were successfully reconstructed, a mandatory additional round of HLA typing was performed to provide reconstructed variant HLA alleles in addition to those available in the IPD-IMGT HLA database (see Figure 1). In this round of HLA typing, NGS reads were aligned not only to the IPD-IMGT HLA database but also to the candidate individualized HLA variant sequences, and the closest matching alleles were assigned accordingly.

[0064] Reconstruction of tumor-specific HLA allele sequences was performed similarly to that guided by germline variants described above. However, in contrast to the germline workflow, all tumor-specific HLA alleles generated were retained as valid potential candidates, since tumors may violate the diploid background assumption (due to somatic copy number alterations (CNAs) affecting the ploidy of HLA genes or tumor heterogeneity) (see Figure 1).

[0065] Allele-specific expression quantification A particularly important optional step to deconvolute ambiguous class II HLA alleles arising due to suboptimal phasing of heterozygous variants is to perform allele-specific expression quantification from RNA-Seq, if possible. Phasing resolution using short-read NGS data was dramatically affected by the long introns and homopolymeric regions that are frequent in class II alleles (because read pairs no longer span this distance adequately). To obtain reliable expression levels of donor HLA alleles, we relied on fine mapping of RNA-Seq reads from donors to specific HLA sequences. To that end, we used Kallisto 48 (v0.43.1) was used for transcript isoform level expression quantification. However, the functionality of Kallisto was extended whereby previously inferred HLA genotypes were re-assigned to their corresponding HLA sequences using RNA-Seq reads as indexes, and then an attempt was made to deconvolute the correct alleles from the expression of the corresponding isoforms of the inferred HLA alleles. Thus, this step serves two purposes, firstly to attempt to confirm the correct HLA when phasing is not complete, and secondly to provide an HLA allelic abundance measurement reported as transcripts per million mapped reads (TPM).

[0066] In silico spike-in of germline variants to simulate de novo HLA allele discovery The ability of the personalized HLA typing step to infer novel or uncharacterized HLA genotypes was first evaluated with simulated novel HLA alleles (summarized here and in Figure 2). To preserve the sequencing error profile and complexity of the biological data, and thus to keep the simulation as faithful to reality as possible, BamSurgeon 49Germline variations were spiked into the WES data using . Three different class I alleles belonging to three different normal WES patients were randomly selected and applied to the simulation (HLA-A*68:01:02:01, HLA-B*51:01:01:01, and HLA-C*03:04:01:01). The ability of our approach to predict novel HLA alleles was then evaluated using the WES with spiked-invariant reads as input. A total of 1800 independent simulation experiments were performed containing a total of 3200 variants including SNPs, insertions, and deletions. A wide range of effects of the resulting proteins were simulated, including missense, synonymous, in-frame, frameshift, stop-gain, and stop-loss variants. Variants were spiked in both individually and in phase co-occurrence with the aim of modeling more dissimilar alleles. The simulation framework first verified whether the spike-in variant was detected by the germline variant caller; if not, the experiment was marked as a miscall. If the variant was called correctly, the variant HLA allele sequence had to be correctly reconstructed and selected as the best matching HLA allele over its reference HLA counterpart. If the reference allele was output in this process, the experiment was marked as an HLA mistype. These simulations were performed with an additional 36 HLA-A, -B, and -C alleles using 8450 spike-in single variants to demonstrate the robustness of novel HLA allele discovery across a wider range of HLA alleles.

[0067] Simulation of somatic variants in HLA allele sequences Following a similar workflow as the novel HLA allele simulation experiments, the performance of detecting somatic variants in HLA alleles was examined with simulated somatic variants in real data. In this case, somatic variants were spiked into tumor WES reads, while normal WES reads were kept unaltered. A total of 740 simulation experiments were performed covering the same three HLA alleles as in the above cases. The simulations included SNVs as well as small insertions and deletions with variant allele frequencies (VAFs) in the range of 0.01–0.5, allowing for the simulation of heterogeneous tumor subclones or sample contamination. Simulation results were analyzed using the state-of-the-art tumor-specific HLA profiling tool POLYSOLVER (v4). 50 The performance of each tool was then evaluated by its ability to detect each simulated somatic variant.

[0068] Standard WES and HLA targeted sequencing of normal PBMC from 10 donors and WES from matched metastatic melanoma patients To evaluate the performance of the proposed solution on clinical samples, the HLA typing pipeline was applied on 10 WES samples from peripheral blood mononuclear cells (PBMCs) of 10 metastatic melanoma donors. In addition, 14 metastatic melanoma samples (from the same metastatic melanoma cohort) were also included. 51 WES was performed on 10 matched normal PBMCs from 10 randomized controlled trials (10 matched normal PBMCs from 10 randomized controlled trials). All study and ethical approvals and permission with written informed consent from all participants were obtained prior to sample collection. Exome enrichment of samples was performed using the Agilent AllExome v5 kit according to the vendor's protocol, and sequencing was performed on an Illumina HiSeq4000 system using paired-end mode at 151bp / read, generating an average of 50 million reads / sample.

[0069] To verify the accuracy of the results, five samples with at least one potential novel class I allele were subjected to targeted HLA sequencing using the NGSgo-MX11-3 HLA targeted amplification kit and analyzed with NGSengine (v2.20) (GenDx, Utrecht, Netherlands). The NGSgo-MX11-3 kit contains amplification primers for 11 loci (including HLA-A, -B, -C, -DRB1, -DQB1, -DPB1, DRB3 / 4 / 5, DQA1, and DPA) and was multiplexed in three tubes to yield HLA locus-specific amplicons, which were then used for HLA typing. To exclude the possibility of analyses performed with different versions of the IPD-IPD-IMGT HLA database, both WES and HLA targeted sequencing analyses were performed using the same version of the IPD-IMGT / HLA database (v.3.41.2).

[0070] Orthogonal validation of somatic HLA variants using RNA-Seq Validation of somatic HLA variants was then attempted using WES data from 10 metastatic melanoma patients by also performing orthogonal validation with available RNA-Seq data for the same tumor samples. Due to the nature of the RNA-Seq data, only variants located in exonic regions could be subject to this validation. Variants discovered in the WES data were considered confirmed when at least one read carrying the alternative allele was found in the RNA-Seq data. If the expression of the allele was switched off or under-expressed, the variant remained unconfirmed due to false-positive WES calls, low-quality RNA-Seq data, or downregulation of expression. Confirming the presence of variants detected at the RNA level (when feasible) significantly reduced the chance of variants being miscalled due to sequencing errors.

[0071] result Simulation of mistyped or novel HLA allele discovery First, the ability of the personalized HLA typing approach to capture potential mistyped or novel HLA alleles from the simulated HLA variants WES data was evaluated. To evaluate the ability of the approach to detect mistyped or potential novel alleles, 1800 independent simulations were performed with the HLA alleles HLA-A*68:01:02:01, HLA-B*51:01:01:01, and HLA-C*03:04:01:01 (see Figure 2). A total of 3200 HLA variants randomly distributed over the entire length of the selected alleles were used in the simulations. Table 1 summarizes the results, with the experiments categorized according to the number of mutated alleles, variant types, and co-occurring simulated variants. Across all experiments, there was an 83% success rate in detecting simulated variants and correctly inferring novel HLA alleles. Considering only experiments where a single mutation was spiked into an allele, success rates of 97%, 93%, and 96% were observed for SNVs, deletions, and insertions, respectively. As expected, a reduction in success rate was observed when simulating co-occurring phase variants in the same allele. This reduction was particularly pronounced when simulating co-occurring indels due to the difficulty of mapping reads due to mismatches associated with multiple indels. For specific HLA genes, success rates of 86%, 86%, and 74% were observed for the A, B, and C alleles, respectively.

[0072] [Table 1]

[0073] To demonstrate the ability of our approach to identify novel alleles across a broader range of HLA alleles, we extended these experiments to further simulations with 8450 spike-in single variants across a diverse range of 36 additional HLA-A, -B, and -C alleles, achieving an overall success rate of 97% (see Table 2).

[0074] [Table 2]

[0075] Overall, the simulations summarized in Tables 1 and 2 suggest the potential of the proposed approach to recover correct HLA genotypes when mistyping occurs from WES-based NGS data and, therefore, also the potential of this approach to identify novel HLA alleles from WES-based NGS data.

[0076] Validation of WES-based HLA typing at the protein-coding sequence level The OncoHLA HLA typing algorithm was then run on the WES data from PBMCs of the 10 donors. For validation purposes, 5 blood samples from the 10 donors with at least one predicted germline variation in an HLA class I allele were also subjected to targeted HLA sequencing from the GenDx NGSgo-MX11-3 kit (see Methods). The outcomes of this high-resolution targeted HLA sequencing validation are listed in Table 3.

[0077] [Table 3]

[0078] At the protein coding sequence level (i.e., first and second fields of resolution), there was 100% overlap of HLA typing between NeoOncoHLA and OncoHLA with WES data and HLA-targeted sequencing. This 100% performance overlap with HLA-targeted sequencing was a validation of both NeoOncoHLA and OncoHLA for the first and second fields of resolution. 19 .

[0079] OncoHLA for HLA typing from WES data without variant calling integration 19There was a decrease in performance in the third and fourth fields, 86.7% and 70% in the third and fourth fields, respectively (see Table 3). This decrease in performance was expected because both WES-based HLA typing solutions have lower coverage of HLA sequences in the IPD-IMGT HLA database for all HLA exons and non-coding sequences, in addition to the medium read depth of WES, compared to targeted NGS. The performance of our WES-based HLA typing solution was significantly improved when using deep or targeted HLA NGS data as input (see Table 3).

[0080] Integration of germline variant calling will enhance personalized HLA typing and enable novel HLA discovery from WES data Personalized HLA typing, through the integration of variant calling (see Figure 1), significantly improved performance at the third and fourth fields of resolution using WES data (see Table 3). Performance was enhanced through the integration of variant calling, demonstrating an increase in accuracy from 86.7% to 96.7% and from 70% to 86.7% at the third and fourth fields, respectively. Table 4 also depicts the differential HLA typing results using WES data versus targeted HLA sequencing data. In all six cases of variant HLA alleles in normal PBMC samples from five patients with at least one germline variant detected, NeoOncoHLA was able to correct the HLA mistype. In one of these cases, the analysis led to the discovery of a novel class I HLA allele that was confirmed by both WES and targeted NGS sequencing and formally designated HLA*B*:44:02:01:52 (see Table 4), thus improving the typing shown in Table 3. Figure 3 illustrates a germline variant found in the 3'UTR of the closest matched reference HLA*B*:44:02:01:03 allele (carrying the novel HLA*B*:44:02:01:52 allele). This class I novel variant was also subjected to transcript level confirmation from a matched tumor sample from the same patient (see Methods: "Allele-specific expression quantification").

[0081] [Table 4]

[0082] Somatic variant calling on class I HLA alleles using individualized germline HLA alleles as reference NeoOncoHLA was then applied to WES data from 14 metastatic melanoma samples from 10 patients. To capture somatic HLA variants with improved fidelity, variants were called using personalized HLA sequences derived from matched normal PBMCs.

[0083] A total of 15 somatic mutations were detected in classical class I alleles across 14 melanoma samples, of which 7 (47%) were found in HLA-A, 5 (33%) in HLA-B, and the remaining 3 (20%) in HLA-C alleles. Using VEP, the functional consequences of somatic variants were inferred from their predicted effect on the resulting protein. In total, 5 (33%) of the 15 variants were located in protein-coding regions, whereas 10 (66%) were in non-coding genomic regions. The predicted functional impact of the 15 detected somatic variants, sorted from low to high impact, was distributed as follows: 5 intronic variants, 2 5'UTR variants, 3 3'UTR variants, 1 synonymous variant, and 4 missense variants. Interestingly, one of the 4 missense variants occurred in the junctional cleft of HLA-A*02:01:01:01.

[0084] The mean variant allele frequency (VAF) across the detected somatic alterations was moderately low (0.13), suggesting broad genomic heterogeneity and the presence of subclonal mutations exclusive to subpopulations of cancer cells. However, one somatic HLA variant located at HLA-A*03:01:01:01 in sample UV1-0009-T01 showed a high VAF of 0.707. This variant was annotated as missense and changes the second amino acid encoded by the allele from alanine to valine. This variant was also confirmed in the tumor RNA-Seq data, and the resulting novel somatic allele was expressed at high abundance (914.75 TPM).

[0085] Orthogonal validation using RNA-Seq data confirmed five somatic HLA variants in protein-coding regions (see Methods). Variants found in WES data were considered confirmed when at least one read carrying the alternative allele was found in the RNA-Seq data. Three of the five exonic variants confirmed in the RNA-Seq data were observed. The combined expression abundance of the four variant HLA allele sequences was calculated using Kallisto. As expected, variant HLA alleles with very low VAF (0.01–0.02) carrying their unconfirmed RNA-seq variants were not expressed. The remaining two alleles had TPMs of 7.23 and 914.76 (the latter being a variant with a high VAF of 0.707 in HLA-A*03:01:01:01 of sample UV1-0009-T01 described above).

[0086] POLYSOLVER applied to the same 14 metastatic melanoma samples detected 28 somatic variants compared to 15 from NeoOncoHLA. Only one variant of the high VAF (0.707) SNV mentioned above was called by both tools. One source of significant difference to note is that NeoOncoHLA uses all alleles available in the IPD-IMGT / HLA database to align NGS reads, including classical class I, classical class II, and non-classical alleles, and remaps and aligns NGS reads to their true origin, whereas POLYSOLVER uses only classical class I alleles. The same variant was reported twice by POLYSOLVER, once per typed HLA allele. This was addressed more strictly in NeoOncoHLA with the filter that multiple variant calling tools must detect somatic variants compared to a single tool in POLYSOLVER, and improved mapping of NGS reads by mapping against class II and non-classical HLA alleles, and also by including an allelic variant phasing step that takes advantage of both WES and RNA-Seq data.

[0087] In the present invention, the discovery of novel HLA alleles is possible through the systematic integration of variant calling tools. Strong validation of HLA typing using database matching methods is demonstrated in the first and second fields of resolution by both in silico simulations and targeted HLA sequencing experiments from the blood of five donors with at least one candidate germline variant in their HLA alleles. This validation performance was consistent with much of the recent literature on the performance of NGS-based HLA typing. The reduced validation performance results observed in the third and fourth fields of resolution were eliminated when applying a germline variant calling strategy. This is particularly promising when ambiguity exists in the fourth field of resolution, given that HLA typing is performed using standard WES data, and thus the coverage of reads in non-coding regions (introns and UTRs) is very sparse. Interestingly, no HLA nonsynonymous variants were ever detected in exons encoding peptide-binding clefts (i.e., exons 2 and 3 for class I and exon 2 for class II) in any of the analyses, reflecting the comprehensive coverage of the first and second fields in the IPD-IMGT HLA database for HLA allelic variants conferring different antigen presentation patterns. Finally, the ability of the described approach to discover new alleles was demonstrated by the detection of a novel allele not yet reported in the IPD-IMGT HLA database belonging to class I (HLA-B). The novel HLA-B allele was given the designation B*44:02:01:52 and has been formally cataloged by the WHO Nomenclature Committee for Factors of the HLA System.

[0088] The approach described herein has the potential to effectively identify novel class I HLA alleles and tumor-specific HLA variants through systematic integration of variant calling. However, the approach is limited to alleles whose sequences share a relatively high degree of similarity with those of known alleles in the HLA reference database. Therefore, the approach was mostly tested with classical class I alleles from the most extensive collection of full-length allele sequences available in the IPD-IMGT HLA database. As more full-length HLA gene sequences are gradually presented, the method may be performed in the future with class II and non-classical alleles. For the identification of highly dissimilar novel HLA alleles (those that exhibit structural variants such as large insertions or deletions) from NGS data, a de novo alignment approach may be more optimal. 52 Additionally, the individualized germline HLA variant calling step of this study was limited to matching of three state-of-the-art variant callers. Numerous NGS-based germline variant callers have been developed in recent years. 53 As solutions for variant detection in complex genomic regions such as HLA continue to emerge, 54、55 To optimize the accuracy of germline variant detection in the HLA region, it may be beneficial to explore ensemble approaches to variant calling, receiving input from a large number of germline variant callers. 56 .

[0089] A well-known source of ambiguous HLA typing results is characterized by the difficulty of consistently and reliably reconstructing well-phased HLA haplotypes. 16The difficulty of this approach can be attributed to the very nature of short reads in WES data. To be properly phased, short NGS reads must properly cover the variant regions, which has been particularly challenging for class II HLA alleles, as the long homozygous stretches common to HLA alleles prevent read pairs from reliably spanning the distance between variants. Inappropriate phasing is currently a vexing problem in the HLA genotyping field and constitutes a major source of erroneous HLA typing results. 16 When ambiguous phasing occurred in the analysis, RNA-Seq was used only to select the reconstructed HLA alleles for RNA-Seq validation. Long-read sequencing could also be used to address the challenge of haplotype phasing for HLA alleles carrying long intronic regions.

[0090] In summary, the described HLA typing approach demonstrated good performance in full-length HLA typing and the ability to detect novel HLA alleles when validated using targeted HLA sequencing. With the large amount of NGS data that continues to accumulate worldwide in many clinical projects, this approach will enable more accurate full typing, leading to the discovery of novel HLA alleles and helping to fill gaps in the IPD-IMGT / HLA database. Furthermore, performing somatic variant calling with individualized germline reference HLA genotypes may allow the identification of tumor-specific HLA variants with increased fidelity and help characterize HLA-associated tumor immune escape.

[0091] References 1. Robinson J, Malik A, Parham P, Bodmer JG, Marsh SG: IMGT / HLA database--a sequence database for the human major histocompatibility complex. Tissue Antigens 2000, 55:280-287. 2. Robinson J, Halliwell JA, Hayhurst JD, Flicek P, Parham P, Marsh SG: The IPD and IMGT / HLA database: allele variant databases. Nucleic Acids Res 2015, 43:D423-431. 3. Horton R, Wilming L, Rand V, Lovering RC, Bruford EA, Khodiyar VK, Lush MJ, Povey S, Talbot CC, Jr., Wright MW, et al: Gene map of the extended human MHC. Nat Rev Genet 2004, 5:889-899. 4. Morishima Y, Sasazuki T, Inoko H, Juji T, Akaza T, Yamamoto K, Ishikawa Y, Kato S, Sao H, Sakamaki H, et al: The clinical significance of human leukocyte antigen (HLA) allele compatibility in patients receiving a marrow transplant from serologically HLA-A, HLA-B, and HLA-DR matched unrelated donors. Blood 2002, 99:4200-4206. 5. Bray RA, Hurley CK, Kamani NR, Woolfrey A, Muller C, Spellman S, Setterholm M, Confer DL: National marrow donor program HLA matching guidelines for unrelated adult donor hematopoietic cell transplants. Biol Blood Marrow Transplant 2008, 14:45-53. 6. Lee SJ, Klein J, Haagenson M, Baxter-Lowe LA, Confer DL, Eapen M, Fernandez-Vina M, Flomenberg N, Horowitz M, Hurley CK, et al: High-resolution donor-recipient HLA matching contributes to the success of unrelated donor marrow transplantation. Blood 2007, 110:4576-4583. 7. Bodmer W, Bonilla C: Common and rare variants in multifactorial susceptibility to common diseases. Nat Genet 2008, 40:695-701. 8. Trowsdale J, Knight JC: Major histocompatibility complex genomics and human disease. Annu Rev Genomics Hum Genet 2013, 14:301-323. 9. Lee MT, Mahasirimongkol S, Zhang Y, Suwankesawong W, Chaikledkaew U, Pavlidis C, Patrinos GP, Chantratita W: Clinical application of pharmacogenomics: the example of HLA-based drug-induced toxicity. Public Health Genomics 2014, 17:248-255. 10. Fan WL, Shiao MS, Hui RC, Su SC, Wang CW, Chang YC, Chung WH: HLA Association with Drug-Induced Adverse Reactions. J Immunol Res 2017, 2017:3186328. 11. Richard C, Fumet JD, Chevrier S, Derangere V, Ledys F, Lagrange A, Favier L, Coudert B, Arnould L, Truntzer C, et al: Exome Analysis Reveals Genomic Markers Associated with Better Efficacy of Nivolumab in Lung Cancer Patients. Clin Cancer Res 2019, 25:957-966. 12. Ivanova M, Shivarov V: HLA typing meets response to immune checkpoint inhibitors prediction: A story just started. Int J Immunogenet 2021, 48:193-200. 13. Chowell D, Krishna C, Pierini F, Makarov V, Rizvi NA, Kuo F, Morris LGT, Riaz N, Lenz TL, Chan TA: Evolutionary divergence of HLA class I genotype impacts efficacy of cancer immunotherapy. Nat Med 2019, 25:1715-1720. 14. Profaizer T, Pole A, Monds C, Delgado JC, Lazar-Molnar E: Clinical utility of next generation sequencing based HLA typing for disease association and pharmacogenetic testing. Hum Immunol 2020, 81:354-360. 15.Bravo-Egana V, Sanders H, Chitnis N: New challenges, new opportunities: Next generation sequencing and its place in the advancement of HLA typing. Hum Immunol 2021. 16. Klasberg S, Surendranath V, Lange V, Schofl G: Bioinformatics Strategies, Challenges, and Opportunities for Next Generation Sequencing-Based HLA Typing. Transfus Med Hemother 2019, 46:312-325. 17. Schofl G, Lang K, Quenzel P, Bohme I, Sauter J, Hofmann JA, Pingel J, Schmidt AH, Lange V: 2.7 million samples typed for HLA by next generation sequencing: lessons learned. BMC Genomics 2017, 18:161. 18. Orenbuch R, Filip I, Comito D, Shaman J, Pe’er I, Rabadan R: arcasHLA: high-resolution HLA typing from RNAseq. Bioinformatics 2020, 36:33-40. 19. Sverchkova A, Anzar I, Stratford R, Clancy T: Improved HLA typing of Class I and Class II alleles from next-generation sequencing data. HLA 2019, 94:504-513. 20. Rotimi CN, Jorde LB: Ancestry and disease in the age of genomic medicine. N Engl J Med 2010, 363:1551-1558. 21. Zhou F, Cao H, Zuo X, Zhang T, Zhang X, Liu X, Xu R, Chen G, Zhang Y, Zheng X, et al: Deep sequencing of the MHC region in the Chinese population contributes to studies of complex disease. Nat Genet 2016, 48:740-746. 22. Flomenberg N, Baxter-Lowe LA, Confer D, Fernandez-Vina M, Filipovich A, Horowitz M, Hurley C, Kollman C, Anasetti C, Noreen H, et al: Impact of HLA class I and class II high-resolution matching on outcomes of unrelated donor bone marrow transplantation: HLA-C mismatching is associated with a strong adverse effect on transplantation outcome. Blood 2004, 104:1923-1930. 23. Mayor NP, Hayhurst JD, Turner TR, Szydlo RM, Shaw BE, Bultitude WP, Sayno JR, Tavarozzi F, Latham K, Anthias C, et al: Recipients Receiving Better HLA-Matched Hematopoietic Cell Transplantation Grafts, Uncovered by a Novel HLA Typing Method, Have Superior Survival: A Retrospective Study. Biol Blood Marrow Transplant 2019, 25:443-450. 24. Vazirabad I, Chhabra S, Nytes J, Mehra V, Narra RK, Szabo A, Jerkins JH, Dhakal B, Hari P, Anderson MW: Direct HLA Genetic Comparisons Identify Highly Matched Unrelated Donor-Recipient Pairs with Improved Transplantation Outcome. Biol Blood Marrow Transplant 2019, 25:921-931. 25. Agarwal RK, Kumari A, Sedai A, Parmar L, Dhanya R, Faulkner L: The Case for High Resolution Extended 6-Loci HLA Typing for Identifying Related Donors in the Indian Subcontinent. Biol Blood Marrow Transplant 2017, 23:1592-1596. 26. Buhler S, Baldomero H, Ferrari-Lacraz S, Nunes JM, Sanchez-Mazas A, Massouridi-Levrat S, Heim D, Halter J, Nair G, Chalandon Y, et al: High-resolution HLA phased haplotype frequencies to predict the success of unrelated donor searches and clinical outcome following hematopoietic stem cell transplantation. Bone Marrow Transplant 2019, 54:1701-1709. 27. Hosomichi K, Shiina T, Tajima A, Inoue I: The impact of next-generation sequencing technologies on HLA research. J Hum Genet 2015, 60:665-673. 28. Carapito R, Radosavljevic M, Bahram S: Next-Generation Sequencing of the HLA locus: Methods and impacts on HLA typing, population genetics and disease association studies. Hum Immunol 2016, 77:1016-1023. 29. Wittig M, Schmoehl M, Koch S, Ziemann M, Goerg S, Lange V, Jacob F, Forster M, Franke A: HLAssign 2.0: An advanced Graphical User Interface for the analysis of short and long read Human Leukocyte Antigen-typing data. bioRxiv 2020:2020.2005.2025.087627. 30. Nordin J, Ameur A, Lindblad-Toh K, Gyllensten U, Meadows JRS: SweHLA: the high confidence HLA typing bio-resource drawn from 1000 Swedish genomes. European Journal of Human Genetics 2020, 28:627-635. 31. Orenbuch R, Filip I, Comito D, Shaman J, Pe’er I, Rabadan R: arcasHLA: high-resolution HLA typing from RNAseq. Bioinformatics 2019, 36:33-40. 32. Lee H, Kingsford C: Kourami: graph-guided assembly for novel human leukocyte antigen allele discovery. Genome Biol 2018, 19:16. 33. Creary LE, Guerra SG, Chong W, Brown CJ, Turner TR, Robinson J, Bultitude WP, Mayor NP, Marsh SGE, Saito K, et al: Next-generation HLA typing of 382 International Histocompatibility Working Group reference B-lymphoblastoid cell lines: Report from the 17th International HLA and Immunogenetics Workshop. Hum Immunol 2019, 80:449-460. 34. Osoegawa K, Vayntrub TA, Wenda S, De Santis D, Barsakis K, Ivanova M, Hsu S, Barone J, Holdsworth R, Diviney M, et al: Quality control project of NGS HLA genotyping for the 17th International HLA and Immunogenetics Workshop. Hum Immunol 2019, 80:228-236. 35. Turner TR, Hayhurst JD, Hayward DR, Bultitude WP, Barker DJ, Robinson J, Madrigal JA, Mayor NP, Marsh SGE: Single molecule real-time DNA sequencing of HLA genes at ultra-high resolution from 126 International HLA and Immunogenetics Workshop cell lines. HLA 2018, 91:88-101. 36. Patterson M, Marschall T, Pisanti N, van Iersel L, Stougie L, Klau GW, Schonhuth A: WhatsHap: Weighted Haplotype Assembly for Future-Generation Sequencing Reads. J Comput Biol 2015, 22:498-509. 37. Spooner W, McLaren W, Slidel T, Finch DK, Butler R, Campbell J, Eghobamien L, Rider D, Kiefer CM, Robinson MJ, et al: Haplosaurus computes protein haplotypes for use in precision drug design. Nat Commun 2018, 9:4128. 38. Wu TD, Reeder J, Lawrence M, Becker G, Brauer MJ: GMAP and GSNAP for Genomic Sequence Alignment: Enhancements to Speed, Accuracy, and Functionality. Methods Mol Biol 2016, 1418:283-334. 39. Tian S, Yan H, Neuhauser C, Slager SL: An analytical workflow for accurate variant discovery in highly divergent regions. BMC Genomics 2016, 17:703. 40. Poplin R, Ruano-Rubio V, DePristo MA, Fennell TJ, Carneiro MO, Van der Auwera GA, Kling DE, Gauthier LD, Levy-Moonshine A, Roazen D, et al: Scaling accurate genetic variant discovery to tens of thousands of samples. bioRxiv 2017. 41. Kim S, Scheffler K, Halpern AL, Bekritsky MA, Noh E, Kallberg M, Chen X, Kim Y, Beyter D, Krusche P, Saunders CT: Strelka2: fast and accurate calling of germline and somatic variants. Nat Methods 2018, 15:591-594. 42. Cibulskis K, Lawrence MS, Carter SL, Sivachenko A, Jaffe D, Sougnez C, Gabriel S, Meyerson M, Lander ES, Getz G: Sensitive detection of somatic point mutations in impure and heterogeneous cancer samples. Nat Biotechnol 2013, 31:213-219. 43. Koboldt DC, Zhang Q, Larson DE, Shen D, McLellan MD, Lin L, Miller CA, Mardis ER, Ding L, Wilson RK: VarScan 2: somatic mutation and copy number alteration discovery in cancer by exome sequencing. Genome Res 2012, 22:568-576. 44. Lai Z, Markovets A, Ahdesmaki M, Chapman B, Hofmann O, McEwen R, Johnson J, Dougherty B, Barrett JC, Dry JR: VarDict: a novel and versatile variant caller for next-generation sequencing in cancer research. Nucleic Acids Res 2016, 44:e108. 45. Larson DE, Harris CC, Chen K, Koboldt DC, Abbott TE, Dooling DJ, Ley TJ, Mardis ER, Wilson RK, Ding L: SomaticSniper: identification of somatic point mutations in whole genome sequencing data. Bioinformatics 2012, 28:311-317. 46. Erik Garrison GM: Haplotype-based variant detection from short-read sequencing. arXiv preprint 2012, 1207.3907. 47. McLaren W, Gil L, Hunt SE, Riat HS, Ritchie GR, Thormann A, Flicek P, Cunningham F: The Ensembl Variant Effect Predictor. Genome Biol 2016, 17:122. 48. Bray NL, Pimentel H, Melsted P, Pachter L: Near-optimal probabilistic RNA-seq quantification. Nat Biotechnol 2016, 34:525-527. 49. Ewing AD, Houlahan KE, Hu Y, Ellrott K, Caloian C, Yamaguchi TN, Bare JC, P’ng C, Waggott D, Sabelnykova VY, et al: Combining tumor genome simulation with crowdsourcing to benchmark somatic single-nucleotide-variant detection. Nat Methods 2015, 12:623-630. 50. Shukla SA, Rooney MS, Rajasagi M, Tiao G, Dixon PM, Lawrence MS, Stevens J, Lane WJ, Dellagatta JL, Steelman S, et al: Comprehensive analysis of cancer-associated somatic mutations in class I HLA genes. Nat Biotechnol 2015, 33:1152-1158. 51. Aamdal E, Inderberg EM, Ellingsen EB, Rasch W, Brunsvig PF, Aamdal S, Heintz KM, Vodak D, Nakken S, Hovig E, et al: Combining a Universal Telomerase Based Cancer Vaccine With Ipilimumab in Patients With Metastatic Melanoma - Five-Year Follow Up of a Phase I / IIa Trial. Front Immunol 2021, 12:663865. 52. Hayashi S, Moriyama T, Yamaguchi R, Mizuno S, Komura M, Miyano S, Nakagawa H, Imoto S: ALPHLARD-NT: Bayesian Method for Human Leukocyte Antigen Typing and Mutation Calling through Simultaneous Analysis of Normal and Tumor Whole-Genome Sequence Data. J Comput Biol 2019, 26:923-937. 53. Koboldt DC: Best practices for variant calling in clinical sequencing. Genome Med 2020, 12:91. 54. Cooke DP, Wedge DC, Lunter G: A unified haplotype-based method for accurate and comprehensive variant calling. Nat Biotechnol 2021. 55. Poplin R, Chang PC, Alexander D, Schwartz S, Colthurst T, Ku A, Newburger D, Dijamco J, Nguyen N, Afshar PT, et al: A universal SNP and small-indel variant caller using deep neural networks. Nat Biotechnol 2018, 36:983-987. 56. Anzar I, Sverchkova A, Stratford R, Clancy T: NeoMutate: an ensemble machine learning framework for the prediction of somatic mutations in cancer. BMC Med Genomics 2019, 12:63.

Claims

1. 1. A method for characterizing the HLA status of a genetic sample obtained from a subject, comprising: i. performing DNA or RNA sequencing on the genetic sample obtained from the subject; ii. aligning the obtained sequences to one or more reference HLA allele sequences; iii. Determining the HLA status by applying variant calling techniques to identify the presence or type of variants in the HLA sequences of the genetic sample; The method includes:

2. The method of claim 1 , wherein the sample is obtained from a tumor-associated tissue of the subject.

3. The method of claim 2, wherein the reference HLA sequence is a reconstituted germline HLA sequence derived from the subject.

4. The method of claim 1 , wherein the sample is obtained from a tissue for transplant.

5. The method of claim 1 , wherein the sample is obtained from the subject's blood.

6. 6. The method of claim 1, claim 4, or claim 5, wherein the reference HLA allele sequences are derived from an HLA reference library.

7. The method of claim 6, wherein the HLA reference library is the IPD-IMGT HLA database.

8. The method according to any one of claims 1 to 7, wherein the sequencing technique used is whole genome sequencing.

9. The method according to any one of claims 1 to 7, wherein the sequencing technique used is exonic sequencing.

10. The method according to any one of claims 1 to 7, wherein the sequencing technique used is Sanger sequencing.

11. The method according to any one of claims 1 to 7, wherein the sequencing technique used is targeted HLA sequencing.

12. The method of any one of claims 1 to 11, wherein the HLA status of class I HLA alleles is determined.

13. The method of any one of claims 1 to 11, wherein the HLA status of class II HLA alleles is determined.

14. The method of any one of claims 1 to 13, wherein step (iii) is carried out using one or more different germline variant calling techniques.

15. 15. The method of claim 14, wherein the germline variant calling has a read depth of >10 and a variant allele frequency of >0.

30.

16. The method of any one of claims 1 to 15, wherein step (iii) is carried out using one or more different somatic variant calling techniques.

17. 17. The method of claim 16, wherein the somatic variant calling has a read depth of >10 and a minimum number of 3 alternate reads.

18. The method of any one of claims 1 to 17, wherein determining the HLA status comprises determining variants in protein coding and / or non-coding regions.

19. 19. The method of any one of claims 1 to 18, further comprising the step of determining the phase relationship or haplotype of heterozygous variants along the two target alleles after the variant calling technique.

20. 20. The method of claim 19, wherein variant HLA gene, transcript and / or protein sequences are reconstructed.

21. The method of any one of claims 1 to 20, wherein allele-specific expression quantification from RNA sequencing is performed.

22. 22. A method for determining suitability for organ transplantation, comprising determining the HLA status of both the recipient and the donor by carrying out a method according to any one of claims 1 to 21, and comparing the HLA status of the recipient and the donor to confirm a suitable HLA match.

23. 22. A method of determining whether a subject will respond to cancer immunotherapy, comprising determining the HLA status of the subject by carrying out a method according to any one of claims 1 to 21, and comparing the HLA status with specific HLA types known to respond to said immunotherapy.

24. The immunotherapy i. checkpoint inhibition therapy; ii. Adoptive Cell Therapy; iii. Therapeutic T-cell cancer vaccines, or combinations thereof 24. The method of claim 23, wherein said compound is selected from the group consisting of

Images