The breast cancer risk SNP analysis provides a working model of the biology for non-protein-coding risk alleles. Both cis- based and trans- based analyses were performed. In the cis- analysis, we observed 3 transcripts associated with 3 risk loci at an FDR<0.1 (2q35/IGFBP5, 5q11/C5orf35, and 16q12/TOX3). Interestingly, the association with TOX3 (in the same direction) has been previously observed further demonstrating the utility of publicly available data (Riaz et al., 2012). IGFBP5 is a strong candidate gene for mediating the effect of the 2q35 risk locus. In breast cancer cell lines, IGFBP5 overexpression affected cell cycle regulation and apoptosis (Butt et al., 2003). In transgenic mice, IGFBP5 expression interferes with mammary epithelial development (Tonner et al., 2002). IGFBP5 expression levels have also been shown to be a marker of poor outcome in patients with breast cancer (Becker et al., 2012).
