TY - JOUR T1 - Quantifying the degree of sharing of genetic and non-genetic causes of gene expression variability across four tissues JF - bioRxiv DO - 10.1101/053355 SP - 053355 AU - Alfonso Buil AU - Ana Viñuela AU - Andrew A. Brown AU - Matthew N. Davies AU - Ismael Padioleau AU - Deborah Bielser AU - Luciana Romano AU - Daniel Glass AU - Paola Di Meglio AU - Kerrin S. Small AU - Timothy D. Spector AU - Emmanouil T. Dermitzakis Y1 - 2016/01/01 UR - http://biorxiv.org/content/early/2016/05/13/053355.abstract N2 - Gene expression can provide biological mechanisms which underlie genetic associations with complex traits and diseases, but often the most relevant tissue for the trait is inaccessible and a proxy is the only alternative. Here, we investigate shared and tissue specific patterns of variability in expression in multiple tissues, to quantify the degree of sharing of causes (genetic or non-genetic) of variability in gene expression among tissues. Using gene expression in ~800 female twins from the TwinsUK cohort in skin, fat, whole blood and lymphoblastoid cell lines (LCLs), we identified 9166 significant cis-eQTLs in fat, 9551 in LCLs, 8731 in skin and 5313 in blood (1% FDR). We observed up to 80% of cis-eQTLs are shared in pairs of tissues. In addition, the cis genetic correlation between tissues is > 90% for 35% of the genes, indicating for these genes a largely tissue-shared component of cis regulation. However, variance components show that cis genetic signals explain only a small fraction of the variation in expression, with from 67–87% of the variance explained by environmental factors, and 53% of the genetic effects occurring in trans. We observe a trans genetic correlation of 0 for all genes except a few which show correlation between fat and skin expression. The environmental effects are also observed to be entirely tissue specific, despite related tissues largely sharing exposures. These results demonstrate that patterns of gene expression are largely tissue specific, strongly supporting the need to study higher order regulatory interactions in the appropriate tissue context with large samples sizes and diversity of environmental contexts. ER -