A Simple Computational Method For Studying Genetic-Environment Interactions

A team of researchers from the Weill Cornell Medicine and Cornell University’s Ithaca campus has built a simplified computational method for analyzing genetic-environment interactions and the way they influence disease risk.

The research recently appeared in The American Journal of Human Genetics.

Co-senior author, Dr. Olivier Elemento explained that our genes and our environment matters a lot but the interaction of the two can increase risk for disease.

Lead author Andrew Marderstein said that typically, studying gene-environment interactions creates a huge computational challenge.

“Genotype-environment interaction can be thought of as the situation where some genotypes are much more sensitive to environmental insults than others,” said co-senior author Dr. Clark.

Why is it difficult?

In a typical test for gene-environment interactions, researchers study millions of data points in a pairwise fashion i.e they assess one genetic variant and its interaction with one environmental factor at a time. This type of analysis can become quite intensive.

New Approach

The current simplified method prioritizes and assesses a smaller number of variants in the genome for gene-environment interactions.

Marderstein conveyed this saying that they condensed a problem with 10 million different genetic variants to only tens of variants in the genome.

The main difference with this new approach is that this method assesses the genetic variants which were more likely to effect the BMI whereas previous methods would look at every single genetic variant that could lead to a change in BMI.

The researchers found that looking for sections of DNA associated with the variance in a human characteristic, called a variance quantitative trait locus or vQTL, enabled them to more readily identify gene-environment interactions. Notably, the vQTLs associated with body mass index were also more likely to be associated with diseases that have large environmental influences.

One more are where in the new computational method might be useful is in finding how a person might respond to a drug based on gene-environment interactions.

Journal Reference:
Andrew R. Marderstein, Emily R. Davenport, Scott Kulm, Cristopher V. Van Hout, Olivier Elemento, Andrew G. Clark. Leveraging phenotypic variability to identify genetic interactions in human phenotypes. The American Journal of Human Genetics, 2021; 108 (1): 49 DOI: 10.1016/j.ajhg.2020.11.016

Press Release: Weill Cornell Medicine