EpiGraphDB: A database and data mining platform for health data science

Liu, Yi; Elsworth, Benjamin; Erola, Pau; Haberland, Valeriia; Hemani, Gibran; Lyon, M.; Zheng, Jie; Gaunt, Tom R.

doi:10.1101/2020.08.01.230193

Cited by 18 publications

(32 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To support the existence of the confounders identified by LHC-MR, we used EpiGraphDB [24,25] to systematically identify those potential confounders. The database provided for each potential confounder of a causal relationship, a causal e↵ect on trait X and Y (r1, and r3 in their notation), the sign of the ratio of which (sign(r 3 /r 1 )) was compared to the sign of the LHC-MR estimated t y /t x values representing the strength of the confounder acting on the two traits.…”

Section: Application To Real Summary Statisticsmentioning

confidence: 99%

See 1 more Smart Citation

Simultaneous estimation of bi-directional causal effects and heritable confounding from GWAS summary statistics

Darrous

Mounier

Kutalik

2020

Preprint

View full text Add to dashboard Cite

Mendelian Randomisation (MR), an increasingly popular method that estimates the causal effects of risk factors on complex human traits, has seen several extensions that relax its basic assumptions. However, most of these extensions suffer from two major limitations; their under-exploitation of genome-wide markers, and sensitivity to the presence of a heritable confounder of the exposure-outcome relationship. To overcome these limitations, we propose a Latent Heritable Confounder MR (LHC-MR) method applicable to association summary statistics, which estimates bi-directional causal effects, direct heritability, and confounder effects while accounting for sample overlap. We demonstrate that LHC-MR outperforms several existing MR methods in a wide range of simulation settings and apply it to summary statistics of 13 complex traits. Besides several concordant results, LHC-MR unravelled new mechanisms (how being diagnosed for certain diseases might lead to improved lifestyle) and revealed potential false positive findings of standard MR methods (apparent causal effect of body mass index on educational attainment may be driven by a strong ignored confounder). Phenome-wide search to identify LHC-implied heritable confounders showed remarkable agreement between the LHC-estimated causal effects of the latent confounder and those for the potentially identified ones. Finally, LHC-MR naturally decomposes genetic correlation to causal effect-driven and confounder-driven contributions, demonstrating that the genetic correlation between systolic blood pressure and diabetes is predominantly confounder-driven.

show abstract

Section: Application To Real Summary Statisticsmentioning

confidence: 99%

“…LHC-MR identified a confounder for 16 trait pairs out of the possible 78. In order to support these findings, we used EpiGraphDB [24,25] to systematically identify those potential confounders.…”

Section: Application To Association Summary Statistics Of Complex Traitsmentioning

confidence: 99%

Simultaneous estimation of bi-directional causal effects and heritable confounding from GWAS summary statistics

Darrous

Mounier

Kutalik

2020

Preprint

View full text Add to dashboard Cite

show abstract

“…as shown in a recent study [69], aiming to identify the mediators of height effect on coronary artery disease). While we adopted a hypothesis-driven approach to investigate potential mediators, in future work, data mining platforms such as EpiGraphDB (epigraphdb.org) [70] may be used to facilitate the identification of novel mediator traits/biomarkers, or candidates for multi-mediator MVMR analyses.…”

Section: Discussionmentioning

confidence: 99%

Deciphering how early life adiposity influences breast cancer risk using Mendelian randomization

Vabistsevits

Smith

Sanderson

et al. 2021

Preprint

View full text Add to dashboard Cite

Studies suggest that adiposity in childhood may reduce the risk of breast cancer in later life. The biological mechanism underlying this effect is unclear but is likely to be independent of body size in adulthood. Using a Mendelian randomization framework, we investigated 18 hypothesised mediators of the protective effect of childhood adiposity on later-life breast cancer, including hormonal, reproductive, physical, and glycaemic traits. Our results indicate that, while most of the hypothesised mediators are affected by childhood body size, only IGF-1, testosterone, age at menarche and age at menopause influenced breast cancer risk. However, accounting for those traits in multivariable Mendelian randomization showed that the protective effect of childhood body size still remained. This suggests either a direct effect of childhood body size on breast cancer risk or mediation via other pathways not considered. Our work presents a framework for the systematic exploration of potential biological mediators of disease in Mendelian randomization analysis.

show abstract

“…The GBMI GWAS summary statistics used in the analyses described here are freely accessible on the GBMI website (https://www.globalbiobankmeta.org/). All our MR estimates and colocalization results (including 11,612 protein-disease signals in European ancestry and 9,858 signals in African ancestry) are freely available to browse, query and download via the EpiGraphDB platform 24 (https://epigraphdb.org/multi-ancestry-pwmr/). An application programming interface (API) documented on the site enables users to programmatically access data from the database.…”

Section: Quantification and Statistical Analysismentioning

confidence: 99%

“…We further estimated the consistency of pQTLs across ancestries, identified potential multi-ancestry and ancestry-specific causal protein-disease pairs, and integrated MR findings with observational and clinical trial evidence 23 to prioritize drug targets. We report our results in an openly accessible database: EpiGraphDB 24 (https://epigraphdb.org/multi-ancestry-pwmr/).…”

Section: Introductionmentioning

confidence: 99%

Proteome-wide Mendelian randomization in global biobank meta-analysis reveals multi-ancestry drug targets for common diseases

Zhao

Rasheed

Nøst

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Proteome-wide Mendelian randomization (MR) shows value in prioritizing drug targets in Europeans, but limited data has made identification of causal proteins in other ancestries challenging. Here we present a multi-ancestry proteome-wide MR analysis pipeline based on cross-population data from the Global Biobank Meta-analysis Initiative (GBMI). We estimated the causal effects of 1,545 proteins on eight complex diseases in up to 32,658 individuals of African ancestries and 1.22 million individuals of European ancestries. We identified 45 and seven protein-disease pairs with MR and genetic colocalization evidence in the two ancestries respectively. 15 protein-disease pairs showed evidence of differential effects between males and females. A multi-ancestry MR comparison identified two protein-disease pairs with MR evidence of an effect in both ancestries, seven pairs with European-specific effects and seven with African-specific effects. Integrating these MR signals with observational and clinical trial evidence, we were able to evaluate the efficacy of one existing drug, identify seven drug repurposing opportunities and predict seven novel effects of proteins on diseases. Our results highlight the value of proteome-wide MR in informing the generalisability of drug targets across ancestries and illustrate the value of multi-cohort and biobank meta-analysis of genetic data for drug development.

show abstract

EpiGraphDB: A database and data mining platform for health data science

Cited by 18 publications

References 39 publications

Simultaneous estimation of bi-directional causal effects and heritable confounding from GWAS summary statistics

Simultaneous estimation of bi-directional causal effects and heritable confounding from GWAS summary statistics

Deciphering how early life adiposity influences breast cancer risk using Mendelian randomization

Proteome-wide Mendelian randomization in global biobank meta-analysis reveals multi-ancestry drug targets for common diseases

Contact Info

Product

Resources

About