Identifying novel genes associated with breast cancer susceptibility using differential allelic expression ratios

• Autor(a) principal: Martins, Catarina Pinto
• Data de Publicação: 2019
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10400.1/12625
• topic: Cancro da mama; Risco; Cis-regulação; OCIAD1; GRHL2

Breast Cancer (BC) is the most common cancer among women worldwide. However, the current knowledge of BC susceptibility only accounts for half of the familial cases. The few functional studies performed for genome-wide association studies (GWAS) loci revealed a role for cis-regulatory variation, suggesting that risk variants may be acting by regulating gene expression levels. Therefore, we hypothesise that the most efficient approach to tackle BC missing heritability is to focus susceptibility studies on variants with greater cis-regulatory potential. Hereby, we present an innovative approach to genetic association studies, using a quantifiable readout of the effect of cis-regulatory variants — differential allelic expression (DAE). To identify candidate risk genes for our study, we selected Single Nucleotide Polymorphisms (SNPs) weakly associated with BC risk in GWAS and in the iCOGS consortium and identified their proxy SNPs. The resulting 591 candidate risk variants were located in 92 different genes, of which 41 had evidence of being cisregulated in a DAE study of normal breast tissue. The clinical impact of these genes was assessed, for a diverse list of clinical variables (differential expression analysis, FDR ⩽ 1% and absolute fold-change ⩾1.5). A final list of 18 risk candidates cis-regulated and with clinical impact genes was identified. OCIAD1 and GRHL2 genes were selected to perform case-control association studies using DAE values. DAE of OCIAD1 was significantly associated with BC risk (p-value=0.002 and 0.008, in two independent experiments using blood samples), while DAE of GRHL2 needs further validation of association (pvalue= 0.014 and 0.096, in two independent experiments in breast tissue). This project proved that association studies using DAE as a quantifiable variable, together with the whole pipeline used to select the candidate genes, is an efficient approach to detect novel risk genes for BC