Study of the association of differential allelic expression with breast cancer clinical features

• Autor(a) principal: Góis, António Manuel Lourenço de
• Data de Publicação: 2023
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10400.1/19682
• topic: Allelic expression (ae) analysis; Breast cancer; Cis-regulation; Tumours clinical features; Domínio/Área Científica::Ciências Médicas::Outras Ciências Médicas

Genome-wide association studies have identified thousands of low-risk variants associated with BC. Since most associated variants are in non-coding regions of the genome and can be in linkage disequilibrium with many others, it’s difficult to pinpoint the true causal variant and its target gene. Post-GWAS functional analysis have shown that altering the expression of genes by cis-acting variants is a common mechanism involved in risk for breast cancer (BC). Therefore, we hypothesize that germline cis-regulatory variants may determine/contribute to tumour clinical features. We propose to test the association of allelic expression (AE) ratios with BC tumours clinical features and patient’s survival. Differences in allelic ratios distributions between tumours with different molecular profiles would indicate a different genetic background for cis-regulatory variants between these tumours. To perform this analysis, we used The Cancer Genome Atlas – Breast Cancer (TCGA-BRCA) normal-matched dataset, and with the R programming language, we complemented this data with allelic expression ratios in various SNPs previously calculated, resulting in a dataset of 105 patients with both clinical and AE data. We identified 63 variants displaying significant effect sizes (|Hedges’s G| ≥ 0.5, 95% interval not crossing 0) for ER statuses, 119 for PR status and 58 for HER2 status, some of them associated with two receptor statuses. Only one variant (rs3764859 in COQ6 and ENTPD5 genes) was found statistically associated with ER status after correction for multiple testing (q-value ≤ 0.1). Additionally, six of the variants were also associated with survival including rs2236225, a variant located at gene MTHFD1 that showed a larger expression difference in both negative ER and PR populations compared to their positive counterparts. We found evidence that cis-acting variants may determine BC clinical features and disease development, and therefore BC subtypes. However, these findings need to be replicated in an independent dataset to be validated.