A landscape genomic approach to investigating growth performance in South African Bonsmara cattle

Abstract

This study utilized a landscape genomic approach to investigate growth performance in Bonsmara cows from three South African provinces, Eastern Cape, Free State and North-West. Landscape genomics seeks to investigate potential associations between the genotype of an animal and a specific environment. Genotype and phenotypic growth trait data from 4679 Bonsmara cattle were obtained for analysis. The cattle were grouped according to province, ownership, population size per province, and sex. After editing and pruning, the final list of animals included 766 cows from the Eastern Cape (418), Free State (224), and North-West (124) provinces. The genotypic data originated from four SNP array panels; GGP 80k (GeneSeek Genomic Profiler™), GGP 150k (GeneSeek Genomic Profiler™), IDB version 3 (International Beef and Dairy), and VersaSNP 50k (Weatherbys Scientific). The common SNPs across these panels were identified and quality control was conducted with PLINK software; 25272 SNPs remained for downstream analysis. The population structure of the cows was analyzed through PCA plots and admixture plots, using GCTA64 and ADMIXTURE software respectively. Weather data for the three provinces included summer and winter month temperatures, relative humidity, and average annual precipitation, from 2016 to 2021. Landscape genomics analysis was conducted on the weather variables and the 25272 common SNPs, using the latent factor mixed model (LFMM) landscape ecology association (LEA) software package in RStudio. The output results consisted of candidate loci that the LEA identified to be associated with the environmental variables included in the weather datasets. The Fst values for each candidate loci list (three lists in total; one per province) were calculated and the 20 SNPs (20 SNPs per candidate loci list; 60 SNPs in total) with the highest values were chosen for gene annotation. The objective of gene annotation was to determine if any of the associated genes were linked to growth performance in cattle. Nine out of the 60 annotated SNPs were found to be associated with genes that had previously been reported in cattle and linked to growth performance. A genome wide association study (GWAS) was also conducted to identify candidate loci associated with eighteen-month weight (18MW) phenotypes. The GWAS was conducted using PLINK and the results were plotted on Manhattan Plots using RStudio software packages. The GWAS results were used comparatively to the candidate loci results from the LEA analysis. There were no commonly associated loci identified between the GWAS and the LEA. Further studies on larger and more informative data sets will be needed for confirmation of the LEA results.