The manatee variational autoencoder model for predicting gene expression alterations caused by transcription factor...

Area of Science:

• Computational Biology
• Systems Biology
• Genomics

Background:

• Understanding how transcription factors (TFs) regulate gene expression is crucial for cell function.
• Predicting the impact of TF perturbations on cellular transcriptomes is a significant challenge in molecular biology.

Purpose of the Study:

• To introduce the Manatee variational autoencoder model for predicting TF perturbation-induced transcriptomes.
• To evaluate the efficacy of Manatee in simulating transcriptomic phenotypes across various cell types.
• To develop a screening approach for identifying TF combinations that induce specific transcriptomic outcomes.

Main Methods:

• Development of the Manatee variational autoencoder model.
• In silico perturbation analysis to predict transcriptome changes.
• In silico screening analysis for TF combination prioritization.

Main Results:

• Manatee accurately recapitulates target transcriptomic phenotypes in diverse cellular lineage transitions.
• The model demonstrates predictive power for TF-induced changes in gene expression.
• In silico screening successfully identified TF combinations for desired transcriptomic phenotypes.

Conclusions:

• The Manatee model provides a powerful tool for predicting the effects of TF perturbations.
• Manatee facilitates the exploration of TF combinations to engineer specific cellular states.
• This approach has implications for understanding and manipulating cellular differentiation and function.