Figure 1

Strategy for candidate gene prioritization in WGAS results. Given a WGAS such as the breast cancer study of the CGEMS initiative [2], ~500,000 SNPs were initially interrogated, which represent a lower number of linkage disequilibrium (LD) blocks in which 24,458 known human genes are distributed. Even when a clear LD block contains several significant SNPs, different genes may be present and molecular and/or functional analyses are required to determine the most likely candidates and their interactions. To obtain this information at the genome-wide level, we propose first to use GO terms to examine the WGAS rank for asymmetries in biological processes. These asymmetries will then be used to guide the analysis of omic data sets relevant to breast cancer biology. Next, higher-level data analyses – protein-protein interactions that may be over-represented for the same processes, and variants in cis/trans affecting germline gene expression levels that lead to hypotheses on the possible functional effects of risk alleles – are performed using a combination of evidences, WGAS results and recognized low-penetrance susceptibility genes/proteins or benchmarks.