1. Home
  2. node
  3. 3521

PQHS 501: Fine-Mapping of Overall and Triple-Negative Breast Cancer Risk Loci by Integrating Multiple Tissue Gene Expression Data

Event Date:
October 24th 9:30 AM - 10:30 AM

Mengxuan Li

Presented by Biomedical & Health Informatics (BHI) PhD trainee Mengxuan Li in Biomedical Research Building room 105

Breast cancer (BC) is a commonly diagnosed cancer in women. Triple-negative breast cancer (TNBC), characterized by the absence of estrogen and progesterone receptors and lack of HER2, is an aggressive subtype and lacks effective targeted therapies. Despite large genome-wide association studies (GWAS) identifying many loci for BC and TNBC risk, the underlying causal variants, target genes and relevant tissues remain largely unknown.

We applied our new algorithm—TGVIS (Tissue-Gene pair, direct causal Variants, and Infinitesimal effect Selector) ((PMC12223092))—to fine-map BC and TNBC GWAS signals. TGVIS searches causal gene-tissue pairs and direct causal variants at a GWAS locus by integrating xQTL data (expression QTLs, alternative polyadenylation QTLs and splicing QTLs) from multiple tissues simultaneously. We obtained European GWAS summary statistics from Breast Cancer Association Consortium for both overall BC and TNBC, and the summary statistics for xQTLs across the 34 tissues from GTEx v10 (N 103-803). African-ancestry female analyses used GWAS data from the African-ancestry Breast Cancer Genetics Consortium and xQTLs from 155 normal female breast tissue samples of African ancestry from the Susan G. Komen Normal Tissue Bank (PMC11065893).

For Europeans, among the 124 BC GWAS loci, TGVIS identified 32 credible sets (CSs) (Pratt index (PI) > 0.15) involving 70 gene–tissue pairs and 14 direct causal variants. Of the 31 identified genes, 29 are supported by breast cancer literature and we confirmed 2 genes as breast genes (CASC10, HNF4G). Breast tissue is the most relevant tissue, followed by lymphocyte and fibroblast based on their frequency among the identified gene-tissue pairs. Similarly, across 12 European TNBC GWAS loci, TGVIS identified 4 CSs involving 6 causal gene–tissue pairs and 1 direct causal variant. All 5 identified genes (NPAT, TOX3, CHD9, ANKLE1, BABAM1) are supported by literature.

For Africans, using female-breast xQTLs, TGVIS identified 6 direct causal variants and 1 gene-tissue pair (HNRNPA1P7-breast) across 7 BC loci with CS-PI >0.15. For the 3 TNBC loci, 1 direct causal variant met the threshold of  CS-PI >0.15, and 1 additional gene–tissue pair (LINC01956–breast) was identified as causal.


If unable to attend in person in Biomedical Research Building room 105, you may join via Zoom at



Meeting ID: 958 2937 2435
Passcode: 087450