Development and Evaluation of Low-density Snp Marker Panels in Indian Cattle Breeds

Abstract

This study aimed to develop and evaluate low-density single nucleotide polymorphism (SNP) marker panels for Indian cattle breeds. The research objectives included creating SNP marker panels using multiple selection criteria and assessing their performance in population structure analysis and genotype imputation. High-density genotyping data from 326 animals representing 15 cattle breeds were retrieved and subjected to quality control measures. Three datasets (A, B, and C) were formulated to include various breed combinations. Low-density SNP panels of five densities (1K, 3K, 5K, 8K, and 10K) were developed using three selection criteria: Delta, Pairwise Wright’s-FST, and Informativeness of Assignment, utilizing TRES software. After preliminary analysis, three breeds were excluded due to suboptimal performance, and 251 animals belonging to 12 cattle breeds were analysed for performance. Performance evaluation was conducted through TRES genomic clustering scores, overlap of criteria, ADMIXTURE clustering, and principal component analysis (PCA). Performance analysis identified Informativeness of Assignment as the best-performing criterion, with the 10K density panel demonstrating superior clustering accuracy for Indian hill, Indian plain, and Exotic cattle cohorts. ADMIXTURE clustering at K=3, K=4, and K=11 confirmed the distinctiveness of clusters for the 10K density panels. PCA further validated the clustering performance of the 10K panels. The inclusiveness of SNPs in the 10K density Informativeness of Assignment panel was assessed based on their distribution across autosomes and protein-coding regions, revealing coverage of all autosomes and association with 36,075 genes in the ARS-UCD1.3 bovine genome assembly. Genotype imputation analysis demonstrated a positive correlation between panel density and imputation accuracy, with the 10K panels achieving the highest accuracy. In conclusion, the 10K density SNP marker panels developed using Informativeness of Assignment criteria were determined to be the most effective for population structure analysis and genotype imputation, providing a better way for genomic studies and breed improvement in Indian cattle.