Publications by authors named "Omer Ronen"
Article Synopsis
- - The text discusses the challenges of detecting complex genetic interactions (epistasis) that influence human traits, pointing out that traditional regression methods struggle with high-order interactions in large genomic datasets due to computational limitations and inadequacies in modeling biological interactions properly.
- - It introduces the epiTree pipeline, built on a framework called Predictability, Computability, Stability (PCS), which utilizes tree-based models to identify higher-order interactions in genomic data by selecting relevant variants based on tissue-specific gene expression and employing iterative random forests.
- - The efficacy of the epiTree pipeline is validated through two case studies from the UK Biobank, demonstrating its ability to reveal both known and novel genetic interactions in predicting traits like red hair and multiple sclerosis, thus potentially
Article Synopsis
- The study challenges the common assumption that genetic variations affect traits in an additive manner by exploring non-additive interactions, specifically in the context of cardiac hypertrophy.
- Researchers used advanced techniques, including low-signal signed iterative random forests and deep learning, to analyze cardiac MRI data from over 29,000 participants in the UK Biobank, revealing complex genetic interactions that traditional methods might overlook.
- The findings highlight a sophisticated gene regulatory network, showing that certain genetic variants interact in intricate ways to influence cardiac structure, pointing to the importance of epistasis in understanding genetic contributions to heart diseases.
The combinatorial effect of genetic variants is often assumed to be additive. Although genetic variation can clearly interact non-additively, methods to uncover epistatic relationships remain in their infancy. We develop low-signal signed iterative random forests to elucidate the complex genetic architecture of cardiac hypertrophy.
View Article and Find Full Text PDF