Panel A: Venn diagrams comparing HLA typing results for HLA-A, HLA-B, HLA-C, HLA-DRB1, HLA-DQA1, and HLA-DQB1 using HIBAG and HLA-LA methods. Each diagram has three sets: total matched (beige), HIBAG (salmon), and HLA-LA (green). The overlaps indicate matches between the methods. Panel B: Venn diagrams comparing HLA-DQ genotypes for tag-SNP, HIBAG, and HLA-LA methods. Each diagram has multiple sets: tag-SNP (salmon), HIBAG (green), HLA-LA (dark blue), tag-SNP and HLA-LA overlap (bright blue), tag-SNP and HIBAG overlap (beige), HLA-LA and HIBAG overlap (sky-blue), and total matched (grayish purple). The overlaps show the consistency between the methods.
HLA and concordance analysis across three HLA typing methods. We then tested three strategies, including HIBAG, HLA*LA, and tagSNPs. We found that the inconsistency arose from HLA-DQA1 imputation. Both HIBAG and HLA*LA make errors in calling HLA-DQA1, particularly DQA1*0501 and DQA1*0505; however, there was a high level of consistency for HLA-DQB1 and DRB1. Combining tagSNPs and HIBAG methods, we were able to impute HLA-DQ genotype with a high degree of consistency, relying principally on the ancestry-specific LD between HLA-DQB1 and DRB1 for the assignment of haplotypes. (A) Venn diagrams of HLA typing results for HLA-LA and HIBAG. Beige: total matched, salmon: HIBAG, green: HLA-LA. (B) Venn diagrams of HLA-DQ genotypes for HIBAG, HLA-LA, and tag-SNP. Salmon: tag-SNP, green: HIBAG, dark blue: HLA-LA, bright blue: tag-SNP and HLA-LA overlap, beige: tag-SNP and HIBAG overlap, sky-blue: HLA-LA and HIBAG overlap, and grayish purple: total matched. HLA-LA: HLA-linear alignment.