A Bayesian Change Point Model for Dynamic Alternative Transcription Start Site Usage During Cellular Differentiation.

Area of Science:

• Molecular Biology
• Genomics
• Bioinformatics

Background:

• Alternative transcription start site (ATSS) usage significantly increases transcript complexity in human tissues.
• ATSS plays a crucial role in cell development, differentiation, and disease, but its dynamic mechanisms are not fully understood.
• Identifying change points in ATSS usage during cell differentiation is vital for understanding these dynamics.

Purpose of the Study:

• To develop a sensitive method for detecting change points in longitudinal ATSS percentage data.
• To analyze differential ATSS events and their associated biological pathways during cell differentiation.

Main Methods:

• Developed a Bayesian change point detection model using reparameterization for beta-distributed percentage data.
• Employed Markov Chain Monte Carlo (MCMC) sampling to obtain posterior distributions of parameters and change points.
• Validated the model's performance through comprehensive simulation studies.
• Applied the model to real data to identify differential ATSS events and perform pathway and transcription factor motif analyses.

Main Results:

• The Bayesian model demonstrated robust and powerful performance across various simulation scenarios.
• The method successfully identified change points in real ATSS longitudinal data.
• Clustering of differential ATSS events based on identified change points revealed distinct patterns.
• Pathway and motif analyses provided biological insights into the functional consequences of ATSS changes during differentiation.

Conclusions:

• The developed Bayesian change point detection model is effective for analyzing ATSS usage dynamics.
• The findings provide crucial biological insights into the mechanisms of cell differentiation driven by ATSS changes.
• This approach enhances the understanding of transcriptomic complexity and its role in biological processes.