Allele-Level KIR Genotyping of More Than a Million Samples: Workflow, Algorithm, and Observations

Abstract

The killer-cell immunoglobulin-like receptor (<i>KIR</i>) genes regulate natural killer cell activity, influencing predisposition to immune mediated disease, and affecting hematopoietic stem cell transplantation (HSCT) outcome. Owing to the complexity of the <i>KIR</i> locus, with extensive gene copy number variation (CNV) and allelic diversity, high-resolution characterization of <i>KIR</i> has so far been applied only to relatively small cohorts. Here, we present a comprehensive high-throughput <i>KIR</i> genotyping approach based on next generation sequencing. Through PCR amplification of specific exons, our approach delivers both copy numbers of the individual genes and allelic information for every <i>KIR</i> gene. Ten-fold replicate analysis of a set of 190 samples revealed a precision of 99.9%. Genotyping of an independent set of 360 samples resulted in an accuracy of more than 99% taking into account consistent copy number prediction. We applied the workflow to genotype 1.8 million stem cell donor registry samples. We report on the observed <i>KIR</i> allele diversity and relative abundance of alleles based on a subset of more than 300,000 samples. Furthermore, we identified more than 2,000 previously unreported <i>KIR</i> variants repeatedly in independent samples, underscoring the large diversity of the <i>KIR</i> region that awaits discovery. This cost-efficient high-resolution <i>KIR</i> genotyping approach is now applied to samples of volunteers registering as potential donors for HSCT. This will facilitate the utilization of <i>KIR</i> as additional selection criterion to improve unrelated donor stem cell transplantation outcome. In addition, the approach may serve studies requiring high-resolution <i>KIR</i> genotyping, like population genetics and disease association studies.