Inferring TF activation order in time series scRNA-Seq studies

Area of Science:

• Computational Biology
• Genomics
• Systems Biology

Background:

• Current single-cell RNA sequencing (scRNA-Seq) analysis methods often neglect transcription factor (TF) information or treat it as a secondary step.
• Integrating TF-gene interactions can enhance model accuracy and provide insights into biological process regulation.

Purpose of the Study:

• To develop a novel computational method, Continuous-State Hidden Markov Models TF (CSHMM-TF), for integrated analysis of scRNA-Seq data and TF regulatory information.
• To identify TFs controlling specific biological paths and determine their order of activation within dynamic cellular processes.

Main Methods:

• Developed the CSHMM-TF method, a probabilistic model that integrates scRNA-Seq data with TF-gene interaction information.
• CSHMM-TF assigns TFs to specific activation points by influencing emission probabilities at later time points in the model.
• Applied and validated CSHMM-TF on multiple mouse and human scRNA-Seq datasets.

Main Results:

• CSHMM-TF successfully identified known and novel TFs regulating the analyzed biological processes.
• The predicted TF activation times align with gene expression data and existing biological knowledge.
• The method demonstrated improved performance over existing methods that do not incorporate TF-gene interaction data.

Conclusions:

• CSHMM-TF provides a robust framework for uncovering TF-driven regulatory dynamics in scRNA-Seq data.
• The method enhances the understanding of gene regulation timing and combinatorial TF control.
• CSHMM-TF represents a significant advancement in the analysis of time-series scRNA-Seq data by leveraging TF information.