Evaluation of HLA-DRB1 imputation using a Finnish dataset

Owing to the vast amount of alleles, high-resolution human leukocyte antigen (HLA) typing is expensive and time-consuming. Scientists have attempted to develop computational approaches to define HLA alleles with high confidence. We tested the reliability of HLA*IMP and SNP2HLA for imputing HLA-DRB1 alleles in a Finnish material (n=161). The per-individual success rates varied between 16.68% and 25.4% using both softwares. One of the most prominent example was HLA-DRB1*01:01 allele showing approximately a 30% success rate while being the most common wrongly imputed allele. In Finland, isolation and migration history have shaped the gene pool narrower showing HLA haplotype frequencies typical to the Finnish population when compared to Europeans. The imputation success for HLA-DRB1 alleles was very low pointing to the importance of population-specific reference material.