17th International HLA Research Articles

Challenges for the standardized reporting of NGS HLA genotyping: Surveying gaps between clinical and research laboratories

Next generation sequencing (NGS) is being applied for HLA typing in research and clinical settings. NGS HLA typing has made it feasible to sequence exons, introns and untranslated regions simultaneously, with significantly reduced labor and reagent cost per sample, rapid turnaround time, and improved HLA genotype accuracy. NGS technologies bring challenges for cost-effective computation, data processing and exchange of NGS-based HLA data. To address these challenges, guidelines and specifications such as Genotype List (GL) String, Minimum Information for Reporting Immunogenomic NGS Genotyping (MIRING), and Histoimmunogenetics Markup Language (HML) were proposed to streamline and standardize reporting of HLA genotypes. As part of the 17th International HLA and Immunogenetics Workshop (IHIW), we implemented standards and systems for HLA genotype reporting that included GL String, MIRING and HML, and found that misunderstanding or misinterpretations of these standards led to inconsistencies in the reporting of NGS HLA genotyping results. This may be due in part to a historical lack of centralized data reporting standards in the histocompatibility and immunogenetics community. We have worked with software and database developers, clinicians and scientists to address these issues in a collaborative fashion as part of the Data Standard Hackathons (DaSH) for NGS. Here we report several categories of challenges to the consistent exchange of NGS HLA genotyping data we have observed. We hope to address these challenges in future DaSH for NGS efforts.

High-resolution HLA allele and haplotype frequencies in several unrelated populations determined by next generation sequencing: 17th International HLA and Immunogenetics Workshop joint report

The primary goal of the unrelated population HLA diversity (UPHD) component of the 17th International HLA and Immunogenetics Workshop was to characterize HLA alleles at maximum allelic-resolution in worldwide populations and re-evaluate patterns of HLA diversity across populations. The UPHD project included HLA genotype and sequence data, generated by various next-generation sequencing methods, from 4,240 individuals collated from 12 different countries. Population data included well-defined large datasets from the USA and smaller samples from Europe, Australia, and Western Asia. Allele and haplotype frequencies varied across populations from distant geographical regions. HLA genetic diversity estimated at 2- and 4-field allelic resolution revealed that diversity at the majority of loci, particularly for European-descent populations, was lower at the 2-field resolution.Several common alleles with identical protein sequences differing only by intronic substitutions were found in distinct haplotypes, revealing a more detailed characterization of linkage between variants within the HLA region. The examination of coding and non-coding nucleotide variation revealed many examples in which almost complete biunivocal relations between common alleles at different loci were observed resulting in higher linkage disequilibrium. Our reference data of HLA profiles characterized at maximum resolution from many populations is useful for anthropological studies, unrelated donor searches, transplantation, and disease association studies.

HLA alleles and haplotypes observed in 263 US families

The 17th International HLA and Immunogenetics Workshop (IHIW) conducted a project entitled “The Study of Haplotypes in Families by NGS HLA”. We investigated the HLA haplotypes of 1017 subjects in 263 nuclear families sourced from five US clinical immunogenetics laboratories, primarily as part of the evaluation of related donor candidates for hematopoietic stem cell and solid organ transplantation. The parents in these families belonged to five broad groups – African (72 parents), Asian (115), European (210), Hispanic (118) and “Other” (11). High-resolution HLA genotypes were generated for each subject using next-generation sequencing (NGS) HLA typing systems. We identified the HLA haplotypes in each family using HaplObserve, software that builds haplotypes in families by reviewing HLA allele segregation from parents to children. We calculated haplotype frequencies within each broad group, by treating the parents in each family as unrelated individuals. We also calculated standard measures of global linkage disequilibrium (LD) and conditional asymmetric LD for each ethnic group, and used untruncated and two-field allele names to investigate LD patterns. Finally we demonstrated the utility of consensus DNA sequences in identifying novel variants, confirming them using HLA allele segregation at the DNA sequence level.

Next-generation HLA typing of 382 International Histocompatibility Working Group reference B-lymphoblastoid cell lines: Report from the 17th International HLA and Immunogenetics Workshop

Extended molecular characterization of HLA genes in the IHWG reference B-lymphoblastoid cell lines (B-LCLs) was one of the major goals for the 17th International HLA and Immunogenetics Workshop (IHIW). Although reference B-LCLs have been examined extensively in previous workshops complete high-resolution typing was not completed for all the classical class I and class II HLA genes. To address this, we conducted a single-blind study where select panels of B-LCL genomic DNA samples were distributed to multiple laboratories for HLA genotyping by next-generation sequencing methods. Identical cell panels comprised of 24 and 346 samples were distributed and typed by at least four laboratories in order to derive accurate consensus HLA genotypes. Overall concordance rates calculated at both 2- and 4-field allele-level resolutions ranged from 90.4% to 100%. Concordance for the class I genes ranged from 91.7 to 100%, whereas concordance for class II genes was variable; the lowest observed at HLA-DRB3 (84.2%). At the maximum allele-resolution 78 B-LCLs were defined as homozygous for all 11 loci. We identified 11 novel exon polymorphisms in the entire cell panel. A comparison of the B-LCLs NGS HLA genotypes with the HLA genotypes catalogued in the IPD-IMGT/HLA Database Cell Repository, revealed an overall allele match at 68.4%. Typing discrepancies between the two datasets were mostly due to the lower-resolution historical typing methods resulting in incomplete HLA genotypes for some samples listed in the IPD-IMGT/HLA Database Cell Repository. Our approach of multiple-laboratory NGS HLA typing of the B-LCLs has provided accurate genotyping data. The data generated by the tremendous collaborative efforts of the 17th IHIW participants is useful for updating the current cell and sequence databases and will be a valuable resource for future studies.

OR24 HLA allele and haplotype frequencies characterized using next-generation sequencing methods in unrelated world-wide populations: Summary from the 17th international HLA and immunogenetics workshop

Aim The primary goal of the 17th IHIWS was to advance the field of Histocompatibility and Immunogenetics research as well as clinical practices through the application of NGS technologies. Analyses of very-high resolution NGS HLA profiles sampled from different populations will be useful for reconstructing human history and may improve match predictions for unrelated donor selection algorithms. Methods In the unrelated population HLA diversity project, 15 laboratories from 10 different countries contributed HLA genotype data generated by various NGS methods. HLA population data included well-defined large datasets from the USA and East Asia and smaller samples from Europe, Australia, and Western Asia. Participants uploaded data in HML/XML format into the 17th IHIW database, data was systematically reviewed, and statistical analysis was performed using the genetic software PyPop v.0.7.0 to estimate; (i) allele frequencies, (ii) heterozygosity, (iii) conformity to HWE, (iv) selective neutrality, (v) 2-locus haplotypes and (vi) LD. The software BIGDAWG was used to estimate extended haplotype frequencies by an EM algorithm. Inter-population variance was determined using Nei’s genetic distances which were then used to construct phylogenetic dendrograms. Results Amongst each population, test of selective neutrality often revealed an excess of heterozygous compared with neutral expectations. A comparison of the HLA genetic diversity estimated at the 4 and 2 fields revealed that diversity at the majority of loci, particularly for populations of European descent, was lower at the 2-field resolution. African descent populations had a lower number of distinct HLA alleles at class I and II loci compared to other groups but exhibited higher levels of heterozygosity. Two-locus haplotype analyses revealed distinct HLA associations at the 4 − field. Analyses of extended haplotype frequencies, estimated both in the presence and absence of DP loci, showed unique haplotypes that were common to each ethnic group although some haplotypes were shared amongst different ethnic groups. Conclusions This workshop project provided a vast amount of new information about HLA genetic diversity at the intronic level and will be a useful resource for anthropological studies, disease association studies and transplantation.

P072NGS characterization of extended HLA haplotypes in Jamaican families from the caribbean bone marrow registry: A study of the 17th international HLA & immunogenetics workshop

Aim Family studies are the gold standard for studying HLA haplotype segregation. This information is critical when studying haplotype associations among populations where few members of the population have been typed, rendering bioinformatics algorithms less reliable. Here, we characterize HLA haplotypes in 5 Jamaican families identified through the Caribbean Bone Marrow Registry. Methods Twenty Jamaican individuals of 5 families were tested by NGS HLA genotypes at loci A, B, C, DRB345, DRB1, DQA1, DQB1, DPA1 & DPB1. We initially attempted full gene sequencing but amplification failed repeatedly due to suboptimal DNA quality from buccal swabs. Typing was performed using Scisco Genomics kit on the MiSeq. Identified haplotypes were compared to haplotype frequency ranking among the 5 major US populations (Caucasians (CAU), African Americans (AFA), Hispanics (HIS), Asian Pacific Islanders (API), and Native Americans (NAM)) using the NMDP Haplostats. Haplostats provides data on 6-locus 2-field resolution, which could give a higher estimate than the 9-locus 3-field haplotypes described here. Results Typing was successful in all individuals and all loci except for failed DPB1 typing on 2 subjects due to low DNA quantity. Analysis of 20 members identified 17 unique haplotypes. The most common haplotype among CAU, HIS, and NAM and 2nd most common in AFA (Haplotype A in Table 1) was observed in 3 families. Another haplotype in 2 families (B) was uncommon among CAU, HIS & AFA (ranked 20466, 23219 & 8463, respectively) and not reported in other populations. Overall observed haplotypes were more common in AFA than other population and 2 haplotypes (C & D) were seen only in AFA. One haplotype (E) was seen only in API (ranked 5660). One haplotype (F) was not reported for any population. Conclusions Although the number of tested families is small, this data suggest that the Jamaican population is unique and diverse, and supports the proposition of establishing a local/regional BMT registry to maximize the chances of identifying matched donors.

Saddlebags: A software interface for submitting full-length HLA allele sequences to the EMBL-ENA nucleotide database.

Submission of full-length HLA allele sequences presents a unique challenge, both for high-throughput sequencing laboratories and smaller diagnostic laboratories. HLA's extensive polymorphism means that accurate representation and annotation of allele sequence is of critical importance, and curators of nucleotide databases must establish submission formats to ensure high-quality data and prevent ambiguities. The IPD-IMGT/HLA database is established as the standard repository for HLA sequences, and it is a major goal of the 17th International HLA and Immunogenetics Workshop to fill the IPD-IMGT/HLA database with full-length HLA sequences. The process of preparing sequence annotation and metadata is cumbersome and error prone, and it is desirable to create a straightforward and concise method of preparing sequence submissions. We introduce Saddlebags, a software tool for rapid generation of HLA (novel) full-length allele sequence submissions. HLA allele sequences are submitted first to EMBL European Nucleotide Archive (EMBL-ENA), and metadata is gathered for subsequent preparation of an IPD-IMGT/HLA formatted submission. Combining these steps into a pipeline reduces effort and minimizes errors for submitting laboratories. This software has been used by Maastricht University Medical Center Transplantation Immunology Laboratory to submit 79 novel alleles to EMBL-ENA, and the tool is freely available for the HLA community.

Collection and storage of HLA NGS genotyping data for the 17th International HLA and Immunogenetics Workshop

For over 50 years, the International HLA and Immunogenetics Workshops (IHIW) have advanced the fields of histocompatibility and immunogenetics (H&I) via community sharing of technology, experience and reagents, and the establishment of ongoing collaborative projects. Held in the fall of 2017, the 17th IHIW focused on the application of next generation sequencing (NGS) technologies for clinical and research goals in the H&I fields. NGS technologies have the potential to allow dramatic insights and advances in these fields, but the scope and sheer quantity of data associated with NGS raise challenges for their analysis, collection, exchange and storage. The 17th IHIW adopted a centralized approach to these issues, and we developed the tools, services and systems to create an effective system for capturing and managing these NGS data. We worked with NGS platform and software developers to define a set of distinct but equivalent NGS typing reports that record NGS data in a uniform fashion. The 17th IHIW database applied our standards, tools and services to collect, validate and store those structured, multi-platform data in an automated fashion. We have created community resources to enable exploration of the vast store of curated sequence and allele-name data in the IPD-IMGT/HLA Database, with the goal of creating a long-term community resource that integrates these curated data with new NGS sequence and polymorphism data, for advanced analyses and applications.

High-throughput next-generation sequencing to genotype six classical HLA loci from 96 donors in a single MiSeq run.

Next generation sequencing (NGS) methods have been established as an efficient approach for HLA typing because unlike traditional Sanger sequencing, they provide unambiguous results at a reasonable cost. We previously developed a multi-locus index method to genotype four HLA loci (A, B, C, and DRB1) on the Illumina MiSeq platform. We have now expanded this method to include two additional loci, HLA-DPB1 and DQB1. Contiguous full-length amplicons from 5'UTR through 3'UTR regions were generated using one long-range PCR reaction per locus for each of the six loci from 96 individuals of different ethnicities. The six amplicons from each donor were pooled, enzymatically fragmented and given a donor-specific index. This approach enabled sequencing of 576 loci from 96 individuals in a single MiSeq run. Donor-specific sequence reads were demultiplexed, and allele calls were generated from FASTQ files using commercially available software. Comparison to HLA genotypes generated from Sanger sequence-based typing (SBT) identified no discordances among any of the alleles analyzed in this study. Importantly, this method was able to resolve 22 DPB1 and 20 DQB1 alleles that were ambiguous with the SBT method. Furthermore, a novel allele in each of these two loci was identified, with the DQB1*05:01:24 allele having a frequency of greater than five percent. This method was subsequently validated against a blinded panel of 22 samples from the 17th International HLA and Immunogenetics Workshop. The flexibility of the method is further highlighted by successful genotyping of eight loci comprising all classical HLA loci for a subset of the samples. We now present a high-throughput, high-resolution, scalable NGS HLA typing method to accurately and efficiently genotype all classical HLA class I and II loci.

P021 Managing NGS HLA genotyping data in the 17th workshop database

Aim The 17th International HLA and Immunogenetics Workshops (IHIW) focuses on the application of next generation sequencing (NGS) technologies for histocompatibility and immunogenomic clinical and research goals. However, the scope and sheer quantity of NGS data raises challenges for their analysis, collection, exchange and storage. The 17th IHIW has adopted a centralized approach to these issues, with the goals of developing tools, services and systems to capture and manage these NGS data, and leveraging the vast store of sequence and allele data in the IPD-IMGT/HLA Database. Methods We worked with NGS platform and software developers to define NGS typing reports that record NGS data in a uniform fashion. The IHIW database and data-storage facility houses typing reports, FASTQ files, subject and specimen data, and pedigrees, and manages the accounts and data-access privileges for 17th IHIW participants. IHIW database tools and services are scripted in the Perl, R or Python languages. Results We defined three distinct typing report formats [Histoimmunogenetics Markup Language (HML), IHIW eXtensible Markup Language (XML), and GenDx XML formats] that record HLA genotyping data and meta-data in equivalent fashion. Our tools convert HML and GenDx XML to IHIW XML, automate and streamline the submission of IHIW XML reports, validate each typing report upon submission, and automate the standardized reporting of novel polymorphisms. All 17th IHIW HLA data are standardized to IPD-IMGT/HLA Database version 3.25.0, with genotype data generated under other database versions converted to 3.25.0. Conclusions The 17th IHIW database applies our standards, tools and services to consume, validate and store structured NGS HLA data generated by GenDx, Illumina, Immucor, Omixon and One Lambda in an automated fashion. The HLA genotyping data and meta-data in these typing reports are validated and standardized for use in 17th IHIW projects, and the database itself will be available for future workshops.

P166 Validation of a next generation sequencing based HLA typing system

Aim Very high resolution typing of HLA genes by Next Generation Sequencing (NGS) methods is becoming mainstream in histocompatibility laboratories for the workup of haematopoietic stem cell and solid organ transplant patients and donors. Determining the full sequence for class I and extended exon coverage for class II genes produces HLA types with far fewer ambiguities than traditional Sanger sequencing based typing (SBT) and other lower resolution typing methods. In addition NGS brings benefits of higher throughput and lower cost. Methods We have validated the Illumina TruSight HLAv2 typing kit for HLA-A, B, C, DRB1/3/4/5, DQA1, DQB1, DPA1 and DPB1. A total of 151 samples were tested following the method described by the vendor. All method steps were carried out manually. Testing comprised 6 batches of 24 samples and 1 batch of 12 samples, which were made up of 73 local DNA samples typed previously by a combination of SBT, SSP and SSO, 54 DNA samples from the UCLA reference panel and 24 samples from the 17th International HLA Workshop Study blinded proficiency panel. Reference types for all eleven loci were not available for all local samples, four digit types were available for all eleven loci for the UCLA samples. Sequence analysis and genotype calling was performed using Illumina Assign for Trusight HLA v2.1 and GenDX NGSEngine. HLA type concordance rate, ambiguity rate and accuracy were examined. Results Total percent concordance between the reference type (n = 130) and genotype determined by Assign was as follows: A: 100; B: 99.2; C: 99.2; DRB1: 97.6.5; DRB345: 98.4; DQA1: 99.2; DQB1: 100; DPA1: 100; and DPB1: 97.6. No allele (1st, 2nd and 3rd field) or genotype ambiguity was found for HLA-A, B, C, DRB345 DQB1 or DPA1. An ambiguous type was determined in 4 cases for DRB1, 3 cases for DQA1 and 29 cases for DPB1. In the case of DRB1 and DPB1 this was due to lack of part of the gene sequence and for DQA1 it was due to poor read depth in those samples. Average read depth across all loci was 223. Repeat rates varied between 1.3% (A & B) and 10.9% (DPA1). Conclusions The Illumina TruSight HLAv2 typing kit provides a suitable first pass high resolution typing system for stem cell and solid organ transplant patients and donors. Accuracy is excellent and the ambiguity levels are very low. We intend to introduce this method for routine HLA typing.

P088 Haplotype analyses of classical HLA genes from families

Aim HLA alleles are observed in specific haplotypes because of linkage disequilibrium between particular alleles. HLA typing using Next Generation Sequencing (NGS) platforms allows generating nearly full-length gene sequences and detecting novel alleles. NGS technologies also permit high-throughput HLA typing for many samples in a cost effective manner. To establish HLA haplotypes from families for the 17th International HLA and Immunogenetics Workshops (IHIW), we developed the Family HLA Haplotype (FamHLAHap) software that compares offspring and parents’ HLA genotypes and builds haplotypes in an automated fashion. To optimize and validate the software, we generated haplotypes from 200 quartet and 1500 trio nuclear families. Methods The quartet families include both parents and two children, and occasional extended family members. A trio contains a child who is diagnosed with multiple sclerosis (MS) and two unaffected parental controls. DNA sequencing library was prepared using MIA FORA NGS kits, and sequenced on either MiSeq or NextSeq platforms. HLA types were obtained using MIA FORA software. The FamHLAHap software reports HLA haplotypes according to the order of genes along the chromosome to facilitate identifying potential meiotic recombination, and also compares to the previously established haplotypes. Results We built haplotypes from a total of 1700 families using the FamHLAHap software. Typing of DPB1 locus by NGS-shotgun sequencing and Sanger often results in ambiguous genotypes because of lack of phasing between exons 2 and 3. In the present study virtually all ambiguities were resolved through the analyses of segregation of sequences. Conclusions We have shown the suitability of the FamHLAHap software to efficiently build haplotypes from NGS data in a large number of families included in the 17th IHIW. Understanding haplotype structures allows delineating evolutionary pathways of the HLA region that include both divergent and consequent evolution. The haplotypes generated from family study provide valuable resources for hematopoietic stem cell donor match prediction, and identification of haplotypes associated with autoimmune diseases.