Deciphering lineage-relevant gene regulatory networks during endoderm formation by InPheRNo-ChIP.

Area of Science:

• Developmental Biology
• Computational Biology
• Genomics

Background:

• Understanding gene regulatory networks (GRNs) is crucial for deciphering human embryogenesis.
• Limited sample availability and biological complexity pose challenges to current GRN inference methods.

Purpose of the Study:

• To develop a computational framework, InPheRNo-ChIP, for reconstructing phenotype-relevant GRNs.
• To integrate multimodal data for enhanced accuracy in GRN inference.

Main Methods:

• Developed InPheRNo-ChIP, a probabilistic graphical model framework.
• Integrated RNA-seq, transcription factor (TF)-specific ChIP-seq, and phenotypic labels.
• Applied the framework to infer GRNs governing human embryonic stem cell differentiation into definitive endoderm.

Main Results:

• InPheRNo-ChIP successfully reconstructed phenotype-relevant GRNs, outperforming existing methods in identifying key endoderm development interactions.
• The method accurately identified regulatory interactions involving endoderm markers FOXA2, SMAD2, and SOX17.
• Benchmarking against a CRISPRi study validated the framework's performance.

Conclusions:

• Incorporating phenotypic context is vital for accurate GRN inference.
• Multimodal data integration, including ChIP-seq, RNA-seq, and phenotypic data, synergistically improves phenotype-relevant GRN reconstruction.
• InPheRNo-ChIP provides a robust approach for studying developmental gene regulation.