The human leukocyte antigen (HLA) locus contains the most polymorphic genes in the human genome. These genes play an important role in immune response and much is already known about their role in autoimmunity and infectious disease. Genotyping of the traditional HLA loci is a very time-consuming process, and established methods are not valid for performing high throughput HLA typing for large sample sets. We apply two strategies to overcome this problem:
- Imputation-based approaches, that were specifically developed for the HLA genes within the major histocompatibility complex (MHC), allow for in silico typing of classical alleles in large sample sets from GWA studies. Due to the strong linkage disequilibrium (LD) across the MHC, alleles across the different classical HLA loci show a strong correlation and SNPs outside the HLA genes can give information on HLA types. At our Institute, we use state-of-the-art imputation tools to predict HLA alleles from GWAS data and to investigate and fine map association signals within the MHC for different diseases.
- To achieve high resolution HLA typings in a high throughput manner, we developed a highly automated method, which employs comprehensive in- solution targeted capturing of the complete classic class I and class II loci in combination with next generation sequencing (NGS). The fully automated analysis we implemented allowed for the accurate characterization of HLA-A (0.99 allele calling rate), HLA-B (0.99), HLA-C (0.99), HLA-DRB1 (0.98), HLA-DQA1 (0.99), HLA-DQB1 (0.99), HLA-DPA1 (0.98) and HLA-DPB1 (0.96). Including possible ambiguities and manual verification allows for exact HLA typing.