Onko-Gene

Critical literature watch in the molecular oncology of solid tumors

FREN

Bioinfo & AI

An encyclopedia of human enhancer-gene regulatory interactions

Summary

This study introduces an extensive encyclopedia of human enhancer-gene regulatory interactions, encompassing over 92 million interactions across 1,458 biosamples and 369 cell types. Researchers developed ENCODE-rE2G, a state-of-the-art predictive model, by integrating various data sources including predictive models, chromatin states, three-dimensional contacts, and large-scale genetic perturbations from the ENCODE Consortium. The model was rigorously benchmarked using data from CRISPR perturbation experiments, expression quantitative trait loci (eQTL), and genome-wide association study (GWAS) variants. This resource offers valuable insights into enhancer networks and significantly improves the analysis of noncoding variants linked to complex diseases.

Critical appraisal

This study marks a significant bioinformatic advancement through the development of ENCODE-rE2G, a novel predictive model for enhancer-gene interactions. The model stands out for its architecture, which integrates multiple data sources (chromatin states, 3D contacts, genetic perturbations) and employs an iterative supervised machine learning approach. The benchmarking is robust, comparing ENCODE-rE2G against other models using a comprehensive dataset including 10,356 CRISPR element-gene pairs, over 30,000 eQTLs, and 569 GWAS variants, demonstrating state-of-the-art performance. Training data originates from the ENCODE Consortium, ensuring large scale and diversity (1,458 biosamples, 369 cell types). Clinically, this model could transform the workflow for interpreting noncoding variants, enabling more precise linkage of GWAS-identified variants to their target genes and relevant cell types. This would facilitate a deeper understanding of complex disease mechanisms and improve prioritization for molecular tumor boards.

The summary and the appraisal were written by a language model from the title and abstract of the original article. A curator decided to publish it. Methodology