Integrating machine learning and functional genomics to study cross-species gene regulatory evolution.

Journal: Current opinion in genetics & development
Published Date:

Abstract

Understanding the genetic basis of phenotypic differences across species has been a longstanding goal of evolutionary biology since Darwin. While a recent proliferation of mammalian genomes has provided an unprecedented inventory of sequence differences between species, the vast majority are in noncoding loci, where it remains challenging to link genetic changes to function. Cis-regulatory elements (CREs) control gene expression via combinatorial, redundant, and context-dependent interactions that are both evolutionarily amenable to change but render their gene regulatory logic difficult to decipher. Recent advances in comparative genomics, functional profiling across species, and high-throughput perturbation assays have begun to catalog cross-species differences in gene expression and CRE function. In parallel, machine learning approaches trained on these data are beginning to predict cis-regulatory activity differences from DNA sequences alone. Here, we highlight recent advances in both experimental and computational strategies to study gene regulatory evolution.

Authors

Keywords

No keywords available for this article.