Inferring transcriptional bursting kinetics from single-cell snapshot data using a generalized telegraph model

Area of Science:

• Molecular Biology
• Systems Biology
• Computational Biology

Background:

• Gene expression exhibits inherent stochasticity due to the burst-like nature of transcription.
• The classical telegraph model (CTM) uses Markovian assumptions to model transcriptional bursting.
• Non-exponential gene-state dwell times are increasingly recognized, necessitating advanced models.

Purpose of the Study:

• To develop an interpretable, non-Markovian mathematical model for transcriptional bursting kinetics.
• To introduce an efficient statistical inference method for estimating burst parameters from single-cell data.
• To compare the performance of the new model against the classical telegraph model.

Main Methods:

• Development of the generalized telegraph model (GTM) allowing arbitrary dwell-time distributions.
• Implementation of an approximate Bayesian computation (ABC) framework for parameter inference.
• Application to synthetic data for validation and to genome-wide data from mouse embryonic fibroblasts.

Main Results:

• The GTM successfully incorporates non-exponential dwell times in gene-state switching.
• The ABC inference method efficiently estimates burst frequency and burst size.
• Genome-wide analysis showed GTM estimates lower burst frequency and higher burst size than CTM.

Conclusions:

• The generalized telegraph model (GTM) offers a more accurate representation of transcriptional bursting.
• The proposed inference method is effective for analyzing static single-cell snapshot data.
• GTM provides a powerful tool for inferring dynamic transcriptional bursting kinetics.