A Bayesian Noisy Logic Model for Inference of Transcription Factor Activity from Single Cell and Bulk Transcriptomic...

Area of Science:

• Computational biology
• Systems biology
• Genomics

Background:

• High-throughput sequencing enables gene expression measurement but direct assessment of regulatory mechanisms like Transcription Factor (TF) activity remains challenging.
• Computational methods are needed to reliably infer TF activity from observable gene expression data.

Approach:

• Developed a noisy Boolean logic Bayesian model for TF activity inference.
• Integrated differential gene expression data with causal graphs.
• Created a flexible framework for incorporating biologically motivated TF-gene regulation logic models.

Key Points:

• Validated the method using simulations and controlled over-expression experiments, demonstrating accurate TF activity identification.
• Applied the model to bulk and single-cell transcriptomics data to study fibroblast phenotypic plasticity.
• Provided user-friendly software packages and a web interface for TF activity querying.

Conclusions:

• The developed Bayesian model offers a robust computational approach for inferring TF activity from gene expression data.
• This method advances the understanding of transcriptional regulation, particularly in complex biological processes like cell plasticity.