Computational protocol: Genome–wide association study of carcass weight in commercial Hanwoo cattle

Similar protocols

Protocol publication

[…] Genome-wide association (GWA) tests were performed using single-locus mixed linear model GWAS (EMMAX) using SNP and Variation Suit v8.5.0 (Golden Helix, Inc., USA www.goldenhelix.com). In GWAS, lack of accounting for population structure may lead to spurious association results []. The EMMAX (Efficient Mixed-Model Association expedited) statistic test approach is proven to outperform the principal analysis and genomic control in accounting for sample structure (stratification and relatedness) []. In our analysis, this technique was used to correct for confounding effects due to subpopulation structure and relatedness between individuals. In the EMMAX, the n×n genetic matrix of identity by state between individuals is computed from the genotype data. The model used can be expressed as:Where y is an n×1 the vector of observed phenotypic values of the animals, X is an n×f matrix of fixed effects including mean SNPs and age (month), β is a q×1 vector representing coefficients of the fixed effects, Z is an n×t relating the instances of the random effect to the carcass weight, β coefficient of fixed effects, u the random animal effects, and e residual effect. The proportion of variance explained (pve) by the effects of the marker was estimated with SNP and Variation Suit v8.5.0 (Golden Helix, Inc., USA www.goldenhelix.com) using the EMMA technique []. Age at slaughter was fitted as fixed effect and sire is included as a covariate. We tested the additive and dominance genotypic models. The level of genome-wide significance was set at 3.28×10−6 (0.05/15,235), corresponding to Bonferroni correction for 15,235 multiple independent tests. The general linear model (GLM) procedure of the Statistical Analysis (V.9.1, SAS Inc., Cary, NC, USA) was used to test for significant difference between the genotypes of the 17 SNPs that are significantly associated (p<0.001) with carcass weight. We further used a haplotype-based approach to identify variants associated with carcass weight. Haplotype blocks were defined according to Gabriel method []. According to this method, SNP pairs to be in strong linkage disequilibrium (LD) if the one-sided upper 95% confidence bound on D′ is 0.98 (that is consistent with no historical recombination) and the lower bound is above 0.7. Haplotype trend regression analysis was performed using SNP and Variation Suit v8.4.4 (Golden Helix, Inc., USA www.goldenhelix.com). Genes within the significant SNPs and their close vicinity were explored from National Center for Biotechnology Information database (http://www.ncbi.nlm.nih.gov/). We have further searched for the QTL regions overlap with the identified candidate regions from bovine QTL database (http://bovinegenome.org/bovineqtl_v2/findQTL.html) and (http://www.animalgenome.org/cgi-bin/QTLdb/BT/index). Gene network analysis provides an important insight into the genetic architecture of complex polygenic traits []. To this end, we carried out gene network analysis using the candidate genes by employing GeneMANIA (www.genemania.org/plugin/) implemented in Cytoscape software 3.4 []. […]

Pipeline specifications

Software tools EMMAX, GeneMANIA
Application GWAS
Organisms Bos taurus