Simulation-guided chemical direct reprogramming informed by temporal cellular conversion processes at the single-cell...

Area of Science:

• Computational biology
• Cellular reprogramming
• Regenerative medicine

Background:

• Direct reprogramming (DR) converts somatic cells to target cells without pluripotency, offering a safer alternative to transcription factor methods.
• Identifying effective small molecules for DR is experimentally challenging and time-consuming.
• Small molecule-induced DR minimizes tumorigenic risks compared to other reprogramming strategies.

Purpose of the Study:

• To develop a computational method, SuperDIRECTEUR, for predicting small molecules that facilitate direct reprogramming.
• To simulate and analyze cellular conversion processes during DR using single-cell temporal transcriptome data.
• To identify stage-specific small molecules that guide cell differentiation towards desired cell types.

Main Methods:

• Simulated RNA velocity analysis of single-cell temporal transcriptome data to model cellular conversion.
• Classification of DR processes into primal, immature, and mature stages.
• Application of a simulated annealing variant to identify small molecule combinations mimicking stage-specific gene expression.

Main Results:

• SuperDIRECTEUR successfully predicted candidate small molecules for converting mouse embryonic fibroblasts to induced neurons.
• The method reproduced experimentally verified molecules and identified functionally related compounds.
• Predicted small molecules targeted distinct pathways in early and late stages, supporting neuronal differentiation and maturation.

Conclusions:

• SuperDIRECTEUR is a powerful computational tool for predicting small molecules for direct reprogramming.
• The method enables stage-specific molecule identification, crucial for efficient and controlled cell fate conversion.
• This approach holds significant potential for advancing regenerative medicine applications.