Enhancing TFEA.ChIP with ENCODE regulatory maps for generalizable transcription factor enrichment

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• Identifying transcription factors (TFs) driving gene expression is crucial in functional genomics.
• Existing methods often rely on motif-based or heuristic approaches, which can lack biological grounding.
• The context-specificity of regulatory element-gene associations presents a significant challenge.

Purpose of the Study:

• To introduce TFEA.ChIP, a ChIP-seq-based tool for TF enrichment analysis.
• To link TF binding profiles with differentially expressed genes via experimentally supported cis-regulatory element (CRE)-gene associations.
• To improve TF identification accuracy through novel filtering strategies.

Main Methods:

• TFEA.ChIP integrates TF binding data (ReMap2022) with regulatory maps (ENCODE's rE2G, CREdb).
• Developed filtering strategies based on confidence scores and recurrence across biosamples to address context-specificity.
• Benchmarked performance on 342 curated gene sets and conducted a case study on hypoxia.

Main Results:

• Recurrence-based filtering significantly improved accuracy compared to GeneHancer and other leading tools (BARTv2.0, Lisa, ChEA3, HOMER).
• Accurate and pathway-specific enrichment of hypoxia-inducible factor-related TFs was demonstrated in a hypoxia case study.
• The updated R/Bioconductor implementation offers user-friendly features like automated workflows and expression-based filtering.

Conclusions:

• TFEA.ChIP provides a biologically grounded and accurate method for inferring transcriptional regulators from gene expression data.
• The tool's enhanced accuracy and user-friendly features facilitate integration into standard RNA-seq analysis pipelines.
• TFEA.ChIP offers a robust and accessible solution for functional genomics research.