Multi-ancestry meta-analyses and fine-mapping identified novel genetic variants and putative susceptibility genes for breast cancer risk
Presenting author: Guochong Jia, Division of Epidemiology, Vanderbilt University Medical Center
Co-authored by:
- Jirong Long, Division of Epidemiology, Department of Medicine, Vanderbilt-Ingram Cancer Center, Vanderbilt University School of Medicine
- Julie R. Palmer, Slone Epidemiology Center, Boston University
- Christopher A. Haiman, Department of Preventive Medicine, Keck School of Medicine, University of Southern California
- Wei Zheng, Division of Epidemiology, Department of Medicine, Vanderbilt-Ingram Cancer Center, Vanderbilt University School of Medicine; on behalf of the African-American Breast Cancer Genetics (AABCG) Consortium
Abstract:
Genome-wide association studies have identified common genetic variants for breast cancer in approximately 200 regions, but causal variants their target genes are largely unknown. We conducted multi-ancestry meta-analyses and fine-mapping analyses using genotype data from > 426,000 females of European, Asian, and African ancestry. For each independent signal, 95% credible sets of credible causal variants (CCVs) were identified using SuSiEX. Putative target genes were identified by cis-eQTL analyses and a modified INQUISIT pipeline integrating genomic functional features. We identified 209 risk regions for breast cancer, including 13 novel risk regions (EGFR locus included). Our fine-mapping analyses identified 332 independent association signals with 5,518 CCVs. A median of six CCVs were identified per signal, much smaller than previous analyses with European-ancestry data alone (median 12 CCVs). These CCVs are significantly enriched in functional regions of transcriptions, enhancers, active transcription start sites, open chromatin regions, and transcription factor binding sites. We identified 248 putative target genes, including 31 genes essential in cell proliferation in breast carcinoma cell lines and 17 breast cancer driver genes. Pathway enrichment analyses found that gene sets for estrogen early response, p53 pathway, PI3K/AKT/mTOR signaling, and TNF-α signaling via NF-κB were significantly enriched (FDR <0.05).