Jorge Ferrer
Most of the most common diseases, such as type 2 diabetes or Alzheimer’s disease, have a well-established genetic predisposition. It is of great interest to identify the genetic variants that predispose individuals to these conditions, because if we knew which genes were altered, this would enable us to develop treatments that target the mechanisms of the disease directly.
The problem is that very often the genetic variants involved in this type of disease do not directly affect a gene, but rather parts of the genome that act as ‘switches’, whose function is to activate genes very selectively in very specific cell types. For example, one of these ‘switches’ (or enhancers) may ensure that a gene important for insulin production is expressed exclusively in the cells that produce insulin. In each cell type, there are tens of thousands of these active switches.
In most cases, it is not known with certainty which switches control each gene. Many studies have succeeded in identifying which genes are controlled by genetic variants of specific switches, or have analysed very large sets of variants using a specific type of genomic data. There is often some uncertainty as to which is the actual target gene. This ENCODE-rE2G study is a collaborative effort by a large consortium that systematically integrates different types of data and AI models to better predict which switches regulate each gene, and does so across a wide range of tissues. It serves as a benchmark to help other studies identify the genetic defects that cause diseases, thereby paving the way for the development of treatments capable of reversing the molecular alterations that cause them.