Dear Editor,
The genotype imputation is an efficient and pivotal approach to estimate the unobserved genotypes in the genomic data from the single nucleotide polymorphism (SNP) genotyping arrays or whole-genome sequencing (WGS).The fine mapping of variants in the genome-wide association study (GWAS) could be critically improved by the imputation with an optimal reference panel based on the population-specific haplotypes.1 The significant progresses of imputation were recently achieved by large high-resolution reference panels,such as the 1KGP3 (1000 Genomes Project Phase 3,n=2504),2 the UK10K (10,000 UK Genome Sequences,n =3781),3 the HRC (Haplotype Reference Consortium,n =32,470),4 TOPMed (Trans-Omics for Precision Medicine,n =97,256)5 and the GAsP (The GenomeAsia 100 K Project,n =1654)6(Supplementary information,Table S1).The multi-ethnic 1KGP3 reference panel is most commonly used for the genetic studies of Asian populations.However,the European ancestry-dominant reference panels exhibited poor performance in the genotype imputation for Chinese and other East Asian populations.7-9 The number of well-imputed variants and imputation accuracy depend on the characteristics of the reference panel dataset,including the sample size,sequencing data quality and population composition.