Causal Inference in Genetic Trio Studies

Abstract

We introduce a method to rigorously draw causal inferences---inferences immune to all possible confounding---from genetic data that include parents and offspring. Causal conclusions are possible with these data because the natural randomness in meiosis can be viewed as a high-dimensional randomized experiment. We make this observation actionable by developing a novel conditional independence test that identifies regions of the genome containing distinct causal variants. The proposed Digital Twin Test compares an observed offspring to carefully constructed synthetic offspring from the same parents in order to determine statistical significance, and it can leverage any black-box multivariate model and additional non-trio genetic data in order to increase power. Crucially, our inferences are based only on a well-established mathematical description of the rearrangement of genetic material during meiosis and make no assumptions about the relationship between the genotypes and phenotypes.

References in corpus (5)

• Second-generation PLINK: rising to the challenge of larger and richer datasets
• Estimating high-dimensional intervention effects from observational data
• Genome-Wide Significance Levels and Weighted Hypothesis Testing
• Rejoinder: "Gene Hunting with Hidden Markov Model Knockoffs"
• Knockoffs with Side Information

Cited by in corpus (12)

• What is a randomization test?
• Fast and Powerful Conditional Randomization Testing via Distillation
• Floodgate: inference for model-free variable importance
• A Power Analysis of the Conditional Randomization Test and Knockoffs
• Searching for consistent associations with a multi-environment knockoff filter
• Controlled Variable Selection from Summary Statistics Only? A Solution via GhostKnockoffs and Penalized Regression
• CAD: Debiasing the Lasso with inaccurate covariate model
• Deploying the Conditional Randomization Test in High Multiplicity Problems
• Hypothesis Testing in Sequentially Sampled Data: AdapRT to Maximize Power Beyond iid Sampling
• Multiple conditional randomization tests for lagged and spillover treatment effects
• Learning to Increase the Power of Conditional Randomization Tests
• Conditional Randomization Rank Test