Using a state-space model with hidden variables to infer transcription factor activities

Area of Science:

• Systems Biology
• Computational Biology
• Genomics

Background:

• Gene regulatory networks (GRNs) involve transcription factors (TFs) regulating gene expression.
• Transcription factor activity (TFA) is crucial but challenging to measure directly.
• Existing models often overlook feedback mechanisms within GRNs.

Purpose of the Study:

• To introduce a novel state-space model (SSM) for inferring TFA from gene expression data.
• To incorporate hidden variables representing regulatory motifs like feedback and auto-regulation.
• To provide a probabilistic framework for simulating GRNs and analyzing TFA.

Main Methods:

• Developed a state-space model (SSM) with hidden variables to represent GRN motifs.
• Simulated gene expression data using SSM to train and validate TFA inference.
• Applied SSM to experimental gene expression data from *Escherichia coli* and *Saccharomyces cerevisiae*.

Main Results:

• Successfully inferred transcription factor activity (TFA) profiles by comparing simulated and inferred data.
• Validated inferred TFA profiles using experimental measurements and literature data.
• Demonstrated SSM's capability in analyzing GRNs with complex regulatory structures (e.g., feed-forward loops).

Conclusions:

• The SSM offers a robust probabilistic framework for simulating gene regulatory networks.
• SSM effectively infers transcription factor activity profiles, complementing existing modeling approaches.
• The model enhances understanding of gene regulation by explicitly considering network feedback and hidden variables.