Causal Transcription Regulatory Network Inference Using Enhancer Activity as a Causal Anchor

Area of Science:

• Genomics
• Systems Biology
• Molecular Biology

Background:

• Mammalian cell types exhibit unique gene expression signatures driven by transcription control.
• Existing methods for inferring gene regulatory networks often identify correlations, not causal relationships.

Purpose of the Study:

• To develop a robust statistical framework for inferring causal gene regulatory networks.
• To leverage enhancer RNA (eRNA) expression as a causal anchor for network inference.
• To generate a comprehensive compendium of transcription factor (TF) targets across diverse human cell types.

Main Methods:

• Developed statistical models and likelihood-ratio tests to infer causal gene regulatory networks.
• Utilized enhancer RNA (eRNA) and transcript expression data from the FANTOM Consortium.
• Incorporated quantitative (dosage-dependent) and binary (on/off) models of transcription factor (TF) activity.
• Analyzed TF target prediction using within-cell-type variation and across-cell-type variation.

Main Results:

• Predicted causal TF targets showed significant overlap with experimentally validated targets.
• The model successfully inferred TF targets in mouse embryonic stem cells, macrophages, and erythroblastic leukaemia.
• Analysis revealed that within-cell-type variability is crucial for predicting cell type-specific TF targets.
• A compendium of high-confidence TF targets across diverse human cell and tissue types was generated.

Conclusions:

• The developed framework accurately infers causal gene regulatory networks using eRNA expression.
• The study highlights the importance of considering TF dosage and within-cell-type variation for precise target prediction.
• The generated compendium provides valuable resources for understanding gene regulation in human cells and tissues.