SCOPE: Localizing fate-decision states and their regulatory drivers in single-cell differentiation

Area of Science:

• Developmental Biology
• Computational Biology
• Genomics

Background:

• Pinpointing cell lineage commitment (branchpoints) and epigenetic priming is challenging.
• Current single-cell sequencing methods lack discrete boundary delineation for lineage mapping.
• Understanding these processes is crucial for mapping the molecular logic of cell fate decisions.

Purpose of the Study:

• To introduce SCOPE (Semi-supervised Conformal Prediction), a novel framework for analyzing single-cell data.
• To rigorously delineate discrete boundaries of cell fate commitment and map lineage specification.
• To uncover epigenetic priming and identify its driving transcription factors.

Main Methods:

• Developed SCOPE, a framework utilizing semi-supervised conformal prediction for high-dimensional single-cell data.
• Formalized cell fate uncertainty using conformal inference to localize lineage specification windows.
• Applied SCOPE to multi-omic data, including simulations, mouse hematopoiesis, human hematopoietic datasets, and human retinogenesis.

Main Results:

• SCOPE transforms single-cell measurements into discrete prediction sets of plausible future cell fates.
• The framework accurately localizes biological windows of multipotent progenitor fate specification.
• SCOPE identified epigenetic priming and its transcription factor drivers by comparing chromatin and transcriptomic data.

Conclusions:

• SCOPE provides a statistically grounded method for localizing cell fate decisions across replicates and modalities.
• It offers a robust tool for identifying the onset of lineage specification in complex developmental systems.
• The framework demonstrates broad applicability and recapitulates known fate specification drivers.