Identification of Causal Risk Factors for Pan-Cancers: a Mendelian Randomization Study
Du, B.; Fan, L.; Tang, C.; Xu, S.; Ge, J.; Shang, X.
Show abstract
BackgroundEvidence from observational studies and clinical trials suggests an association between plasma protein and metabolite levels and cancers. However, the causal relationship between them is still unclear. MethodsWe collected genome-wide association study (GWAS) summary statistics of plasma protein levels from the UK Biobank Pharma Proteomics Project (UKB-PPP, 9,216 to 34,090 participants) and plasma metabolites from the GWAS Catalog (3,441 to 8,299 participants), paired with summary statistics of 99 types of cancers from FinnGen database (131,348 to 412,181 participants). We conducted univariable and multivariable Mendelian randomization (MR) analyses to explore the causal association between plasma protein and metabolites and cancers. ResultsWe identified 175 plasma proteins and 28 metabolites causally associated with cancers (p < 1 x 10-5). Notably, BTN2A1 is causally associated with an increased risk of bone and articular cartilage cancer (OR = 1.776, 95% CI = 1.429 - 2.207), colorectal cancer (OR = 1.200, 95% CI = 1.129 - 1.275), eye and adnexa cancer (OR = 2.686, 95% CI = 1.943 - 3.714), lip cancer (OR = 3.004, 95% CI = 2.193 - 4.114), oral cancer (OR = 1.905, 95% CI = 1.577 - 2.302), ovary cancer (OR = 1.265, 95% CI = 1.143 - 1.400), and rectum cancer (OR = 1.393, 95% CI = 1.263 - 1.536). N6- carbamoylthreonyladenosine level is causally associated with various cancers including colorectal cancer (OR = 1.800, 95% CI = 1.444 - 2.243), head and neck cancer (OR = 2.423, 95% CI = 1.665 - 3.525), hepatocellular carcinoma (OR = 6.476, 95% CI = 2.841 - 14.762), oral cancer and skin cancer (OR = 1.271, 95% CI = 1.161 - 1.392). Additionally, all results are available at the online database (www.causal-risk.net). ConclusionsOur MR analysis reveals causal risk factors for cancers.
Matching journals
The top 4 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Serum Lipidome Profiling Reveals a Distinct Signature of Ovarian Cancer in Korean Women 93%
- Red meat intake interacts with a TGF-β-pathway-based polygenic risk score to impact colorectal cancer risk: Application of a novel approach for polygenic risk score construction. 93%
- Cancer incidence and mortality estimates in Arab countries in 2018: A GLOBOCAN data analysis 92%
Similar papers in this journal
- Genomic alterations and abnormal expression of APE2 in multiple cancers 94%
- Multi-omic signatures identify pan-cancer classes of tumors beyond tissue of origin. 93%
- Proteogenomic Characterization of Triple-Negative Apocrine Carcinoma Reveals Molecular Features of Progression and Chemotherapy Response 93%
Similar papers in this journal
- Metabolomic markers of colorectal tumor with different clinicopathological features 92%
- BNIP3 upregulation characterizes cancer cell subpopulation with increased fitness and proliferation 92%
- Pan-cancer molecular signatures connecting aspartate transaminase (AST) to cancer prognosis, metabolic and immune signatures 92%
Similar papers in this journal
- Associations of plasma omega-6 and omega-3 fatty acids with overall and 19 site-specific cancers: a population-based cohort study in UK Biobank 95%
- Long-Term Impact of Stressful Life Events on Breast Cancer Risk: A 36-Year Genetically Informed Prospective Study in the Finnish Twin Cohort 93%
- Methylation scores for smoking, alcohol consumption, and body mass index and risk of seven types of cancer 93%
Similar papers in this journal
- Peripheral blood mononuclear cells exhibit increased mitochondrial respiration after adjuvant chemo- and radiotherapy for early breast cancer 93%
- Comprehensive analysis of potential immunotherapy genomic biomarkers in 1,000 Chinese patients with cancer 92%
- Circulating serum miRNAs predict response to platinum chemotherapy in high-grade serous ovarian cancer 92%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.