To benchmark the imputation accuracy of our pipeline against the previously published imputation method, we performed the KIR gene copy number imputation by using KIR∗IMP software (http://imp.science.unimelb.edu.au/kir/documentation).13 (link) We used phased haplotypes of (a) SNP genotype data (OmniExpressExome BeadChip array) from IHWG dataset and (b) mock GWAS genotype on the Illumina OmniExpressExome BeadChip array constructed from WGS of part of Japanese individuals in KIR reference panel. The haplotype phasing was performed using SHAPEIT2 software. We then uploaded the haplotypes to the software webserver, and obtained the result of imputed KIR gene copy numbers. The imputation accuracy was assessed by comparing the imputed copy number with those inferred from the NGS-based pipeline.
Sakaue S., Hosomichi K., Hirata J., Nakaoka H., Yamazaki K., Yawata M., Yawata N., Naito T., Umeno J., Kawaguchi T., Matsui T., Motoya S., Suzuki Y., Inoko H., Tajima A., Morisaki T., Matsuda K., Kamatani Y., Yamamoto K., Inoue I, & Okada Y. (2022). Decoding the diversity of killer immunoglobulin-like receptors by deep sequencing and a high-resolution imputation method. Cell Genomics, 2(3), 100101.
+ Open protocol