Modeling transcriptional regulation using gene regulatory networks based on multi-omics data sources

Area of Science:

• Genomics
• Systems Biology
• Computational Biology

Background:

• Transcriptional regulation involves complex cis and trans mechanisms, with disruptions linked to disease.
• Previous models focused on cis-regulatory factors, overlooking local chromatin looping and trans-acting influences.
• Integrating both local and distant regulatory features can enhance gene expression prediction accuracy.

Purpose of the Study:

• To develop a comprehensive computational framework for modeling gene expression.
• To incorporate both cis and trans gene regulatory features, including chromatin looping, into predictive models.
• To improve the accuracy and interpretability of transcription factor influence on gene expression.

Main Methods:

• Developed a computational framework weighting transcription factor data using multi-omics gene regulatory networks.
• Incorporated cis and trans acting mechanisms and local chromatin context into prediction models.
• Integrated long-distance chromatin interactions (chromatin looping) between distal binding sites and gene promoters.

Main Results:

• The developed prediction models significantly outperformed models using only cis-regulatory features.
• Models integrating chromatin looping demonstrated further improvements in predictive accuracy.
• Effect estimates from the models were successfully applied to analyze cis-regulatory rare variants in gene expression.

Conclusions:

• The study provides refined effect estimates for individual transcription factors, aiding genome-wide role characterization.
• The framework enables the integration of diverse data types for a unified model of transcriptional regulation.
• This approach enhances understanding of gene expression regulation and its disease implications.