Imputation-based HLA typing with SNPs in GWAS studies


Email: Dr. Xiuwen Zheng
This page was last updated on Aug 30, 2017


Introduction

SNP-based imputation approaches for human leukocyte antigen (HLA) typing take advantage of the extended haplotype structure within the major histocompatibility complex (MHC) to predict classical HLA alleles using dense SNP genotypes, such as those available on chip panels of genome-wide association study (GWAS). These methods enable HLA analyses of classical alleles on existing SNP datasets genotyped in GWAS studies at no extra cost. Here, I describe the workflow of HIBAG, an imputation method with attribute bagging, for obtaining a sample’s HLA class I and II genotypes of two-field resolution using SNP data. Two examples are provided to illustrate with a publicly available HLA and SNP dataset: genotype imputation with pre-fit classifiers in GWAS, and model training to build a new classifier.

Download

HIBAG package: http://www.bioconductor.org/packages/HIBAG
HIBAG.gpu package: https://github.com/zhengxwen/HIBAG.gpu
ImmPuteDataPackage.zip: download (or download from http://immpute-project.immunogenomics.org)
ImmunoChip-Broad-HLARES-HLA4-hg19.RData: download (or download from http://zhengxwen.github.io/HIBAG/platforms.html)

Tutorial

R markdown / HTML

Citation

If you use HIBAG in a published analysis, please report the HIBAG kernel version and cite the appropriate publication or publications listed below: