We estimated nongenetic contributions in gene expression measures using PC analysis (Leek and Storey 2007; Stegle et al. 2010). PCs were estimated from the gene expression values, ignoring the family relationship in the data. Top PCs were included in the eQTL regression model as covariates. In order to determine the optimal number of PCs to use, we used genotyped SNPs and, starting with the first PC, included other PCs one at a time until the number of transcripts mapped locally no longer increased. Finally, the top 69 PCs were used in the Affymetrix expression data set (MRCA), and the top 61 PCs were used for the Illumina expression data set (MRCE).
