Computational protocol: XVIth QTLMAS: simulated dataset and comparative analysis of submitted results for QTL mapping and genomic evaluation

Similar protocols

Protocol publication

[…] Various methods for detecting genome regions affecting single trait were proposed. Karacaoren [] suggested a GWA by ridge regression on actual YD (RR_YD) or YD adjusted by a principal components correction for LD structure (RR_YDadj). Riggio and Pong-Wong [] performed a regional heritability mapping (RHM) approach by fitting a mixed model where the effect of a genomic region (20 SNPs) and the overall genetic background were added as random. Minozzi et al. [] used a selection analysis implemented in the randomForest package and a GWA based on Mixed Model and Regression - Genomic Control (GRAMMAR-GC). Both methods were applied to single (RF_ST; GRM_ST) and multiple trait (RF_MT; GRM_MT) estimated breeding values, and on yield deviations (RF_YD; GRM_YD). Moioli et al. [] applied a selective genotyping (SG) technique based on the comparison of the allele frequencies of sliding windows of 5 consecutive SNPs in two groups divergent for production. Grosse-Brinkhaus et al. [] performed the GWA with the Genome wide rapid association using Mixed Model and Regression (GRAMMAR). Demeure et al. [] applied a linkage analysis (LA) based on a within-half-sibs family linear regression by using QTLmap software. García-Gámez et al. [] performed a GWA based on a mixed model including the pedigree information as a random polygenic effect and each SNP genotype as a fixed effect with the DMU software package. Moreover, they performed a LDLA analysis where, in addition to the polygenic effect, the QTL was fitted as a random effect with a (co)variance structure that was a function of identity-by-descent (IBD) probabilities between haplotypes.Three authors investigated the pleiotropic nature of the QTLs using different approaches. Karacaoen [] performed RR on the two first principal components (PC) extracted from YD and YD adjusted for LD structure. Grosse-Brinkhaus et al. [] proposed two different approaches. First they performed the GRAMMAR procedure on two principal components extracted from the three traits and, as an alternative, they analyzed the data set with a Bayesian multivariate method implemented in the snptest package. Finally, Riggio and Pong-Wong [] estimated correlations between regional EBVs to evaluate possible pleiotropic effects among traits, when a QTL was found significant for more than one trait. […]

Pipeline specifications

Software tools randomforest, QTLMap, SNPTEST
Applications Miscellaneous, WGS analysis, GWAS
Organisms Homo sapiens