A Study of Causal Inference and its Implications in Kinship Correction

Document Type

Article

Publication Date

Summer 2019

JAX Location

In: Student Reports, Summer 2019, The Jackson Laboratory

Abstract

Causal inference is a powerful tool for describing cause and effect relationships and what they mean. These tools and methods originated from artificial intelligence (Al) research and the need for a better understanding of the logic behind cause and effect in order to teach a computer to make the same deductions humans make every day. Causal inference has tools that make it possible to make causal statements that previously could only be made by conducting a randomized control trial. Causal inference also provides an intuitive way to look at confounding variables and correct for their effects. This work aims to ground the process of kinship correction in the language of causal models. The current state-of-the-art method for kinship correction is the efficient mixed-model association (EMMA) statistical test. We will show how causal models and EMMA can be used to understand and correct for the influence of kinship in genetic experiments [1].

Please contact the Joan Staats Library for information regarding this document.

COinS