Back

Simulation and Sensitivity Analysis and Cross-Validation, Demonstrating the Utility of Genteract GxE Discovery methods

Holohan, B. C.; Laderman, R.

2020-12-28 genomics
10.1101/2020.11.25.396861 bioRxiv
Show abstract

Gene-environment interactions are at the heart of why many complex traits are not fully heritable, and why prediction of disease incidence and individual response to environmental changes based on genetics has been underwhelming in utility. Understanding these interactions is the primary limiting factor for the application of personalized medicine, but current methods are not well suited for dealing with complex traits that pose both a dimensionality and sparse data problem to unsupervised analysis methods. Genteract has developed a proprietary analytical technique that allows for detection and interpretation of GxEs regarding specific pairs of a single phenotype with a single environmental factor; these methods allow us to develop a platform that can be used to predict how individuals will respond to changes in their environment based on their genetics. To validate the methods we performed two types of testing: cross-validation against a dataset of clinical study results, and application of the methods in a simulated dataset. These tests enable a greater understanding of the methods utility, statistical power and predictive capabilities.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"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.