Evaluation on genotype imputation performance of three porcine 50K SNP chips from chip data to sequencing data

Objective  Porcine 50K SNP (single nucleotide polymorphisms) chips have been widely used in pig genomic breeding. Meanwhile, genotype imputation can significantly increase the amount of genotype data without increasing the cost of sequencing, which facilitates genetic resolution and genetic evaluation of complex traits. This study was aimed to evaluate the genotype imputation performance from genotype to sequence data of three porcine SNP chips.

Method  A total of 48 Duroc pigs with three kinds of porcine SNP chips were used as target panel to evaluate the genotype imputation accuracy. A total of 260 pigs with whole genome sequencing data formed a reference panel for genotype imputation. The genotype imputation was performed using Beagle5.1 software to compare the imputation effect of Geneseek 50K, ZhongxinⅠ 50K and Liquid 50K.

Result  The numbers of original SNPs in three kinds of chips were 50697, 57466 and 50885 respectively. The imputation accuracies (genotype consistencies) were 0.886, 0.886 and 0.898 respectively after imputation without any quality control. After filtering the imputed SNPs with low reliability DR² (Dosage R-squared) <0.95, the imputation accuracies (genotype consistencies) of three kinds of chips were up to 0.974, 0.976 and 0.969 respectively, and the numbers of remaining SNPs were 3393066, 3139095 and 3320627 respectively.

Conclusion  Genotype data from the three types of porcine SNP chips can be imputed to sequence data with a high imputation accuracy. This study provides useful reference for subsequent breeding application research.