FIERCE: reconstructing dynamic trajectories from the differentiation potency of single cells

Area of Science:

• Computational Biology
• Genomics
• Developmental Biology

Background:

• Single-cell RNA sequencing (scRNA-seq) enables reconstruction of dynamic cellular processes.
• Existing computational methods rely on assumptions about developmental paths.
• These assumptions limit applications in complex biological systems.

Purpose of the Study:

• Introduce FIERCE (Framework for InfERence of veloCity of the Entropy), a novel computational pipeline.
• Predict changes in single-cell differentiation potency during dynamic processes.
• Enable unsupervised inference of cell lineages.

Main Methods:

• FIERCE utilizes a fully unsupervised approach.
• It infers cell lineages directly on the differentiation landscape.
• No prior specification of developmental parameters is required.

Main Results:

• FIERCE successfully reconstructs three well-known mouse differentiation systems.
• The pipeline's accuracy is quantified on simulated data.
• Demonstrates efficacy in predicting cell differentiation potency.

Conclusions:

• FIERCE overcomes limitations of existing methods for reconstructing dynamic cellular processes.
• The framework offers a powerful tool for analyzing complex biological systems.
• Facilitates a deeper understanding of cell lineage inference and differentiation potency.