A Gene-Set Enrichment and Protein–Protein Interaction Network-Based GWAS with Regulatory SNPs Identifies Candidate Genes and Pathways Associated with Carcass Traits in Hanwoo Cattle

Abstract

Non-synonymous SNPs and protein coding SNPs within the promoter region of genes (regulatory SNPs) might have a significant effect on carcass traits. Imputed sequence level data of 10,215 Hanwoo bulls, annotated and filtered to include only regulatory SNPs (450,062 SNPs), were used in a genome-wide association study (GWAS) to identify loci associated with backfat thickness (BFT), carcass weight (CWT), eye muscle area (EMA), and marbling score (MS). A total of 15, 176, and 1 SNPs were found to be significantly associated (p < 1.11 × 10 −7 ) with BFT, CWT, and EMA, respectively. The significant loci were BTA4 (CWT), BTA6 (CWT), BTA14 (CWT and EMA), and BTA19 (BFT). BayesR estimated that 1.1%~1.9% of the SNPs contributed to more than 0.01% of the phenotypic variance. So, the GWAS was complemented by a gene-set enrichment (GSEA) and protein-protein interaction network (PPIN) analysis in identifying the pathways affecting carcass traits. At p < 0.005 (~2,261 SNPs), 25 GO and 18 KEGG categories, including calcium signaling, cell proliferation, and folate biosynthesis, were found to be enriched through GSEA. The PPIN analysis showed enrichment for 81 candidate genes involved in various pathways, including the PI3K-AKT, calcium, and FoxO signaling pathways. Our finding provides insight into the effects of regulatory SNPs on carcass traits.Genes 2020, 11, 316 2 of 22 and eye muscle area (EMA) [4]. Though substantial improvement in carcass and meat quality have been achieved, due to market requirement for higher quality, and for improving the economic value of Hanwoo, continuous improvement of economically important trait is required [5,6]. A genome-wide association study (GWAS) is an affordable and powerful tool to discover candidate genes and loci associated with quantitative traits [7]. GWASs in livestock, including in Hanwoo [8][9][10][11], have resulted in remarkable insights into the genetic architecture of carcass traits. Genetic variation in complex traits such as carcass and meat quality traits are, however, due to the contribution of many mutations with small effects [1,12] (polygenic effect). Though some of these mutations have been successfully identified through GWASs, the high significance thresholds required to correct for the multiple testing problem results in the identification of only SNPs with a large effect size [7,12]. Further, a GWAS does not make use of the fact that genes work together in a network, and multi-allelic QTL might not be captured due to the bi-allelic nature of SNPs [13]. Moreover, epistasis is an important genetic component underlying phenotypic variation that also accounts for missing heritability [14]. Therefore, a GWAS alone might result in only limited understanding of the nature of complex traits [13]. Suggested solutions to overcome this limitation and understand the genetic complexities regulating complex traits are to complement a GWAS with gene-set enrichment, a protein-protein interaction network (PPIN), and pathway analyses [15][16][17][18]. In GSEA and pathway analysis, a g...

Identification of Candidate Genes and Pathways Associated with Obesity-Related Traits in Canines via Gene-Set Enrichment and Pathway-Based GWAS Analysis

“…The samples were genotyped on an Illumina CanineHD BeadChip (Illumina, San Diego, CA, USA) array, which contains 173,662 SNPs. Quality control (QC) was carried out with PLINK v.1.9 software [24] under the following criteria: minor allele frequency <5%, low genotyping call rate <90%, missing genotype calls >10%, Hardy-Weinberg equilibrium at p < 0.000001. The final genotyping call rate was 98.5%.…”

“…A Bonferroni-corrected threshold was applied to correct for multiple testing, and the genome-wide significance threshold was p < 3.74 × 10 −7 (~0.05/135,553). Manhattan and quantile-quantile (Q-Q) plots were generated using CMplot package in R [24].…”

“…However, performing GWAS powerful enough is relatively difficult because of a few factors like small effect sizes of variants in complex disease, and required stringent statistics because of multiple testing problems. As a result, GWAS identifies only large effect sizes of variants [24]. In addition, GWAS does not capture multi-allelic QTL because of the bi-allelic nature of single nucleotide variants (SNVs) and does not account the fact that disease-associated genes or their products work together in a network [22][23][24].…”

“…As a result, GWAS identifies only large effect sizes of variants [24]. In addition, GWAS does not capture multi-allelic QTL because of the bi-allelic nature of single nucleotide variants (SNVs) and does not account the fact that disease-associated genes or their products work together in a network [22][23][24]. Therefore, a robust GWAS analysis with the aim of detailed biological understanding is challenging.…”

“…Therefore, a robust GWAS analysis with the aim of detailed biological understanding is challenging. The solutions proposed to overcome these limitations are to focus on post-GWAS analyses such as functional gene and pathway enrichment analysis, protein-protein interaction network analysis, and translational medicine [22,[24][25][26].…”