Abstract
Cell dynamics and biological function are governed by changing patterns of gene expression. Intricate gene interaction networks orchestrate these changes. Inferring these interactions from data is a notoriously difficult inverse problem. The majority of existing network inference methods work at the population level. They construct static representations of gene regulatory networks, and they do not naturally allow us to infer differences in gene regulation across heterogeneous cell populations. Here we build upon recent dynamical inference methods that model single cell dynamics using Markov processes, which leads to an information-theoretic approach, locaTE, which employs the localised transfer entropy to infer cell-specific, causal gene regulatory networks. LocaTE uses high-resolution estimates of dynamics and geometry of the cellular gene expression manifold to inform inference of regulatory interactions. We find that this approach is superior to static inference methods, often by a significant margin. We demonstrate that factor analysis can give detailed insights into the inferred cell-specific GRNs. In application to three experimental datasets, we demonstrate superior performance and additional insights compared to stancard static GRN inference methods. For example, we recover key transcription factors and regulatory interactions driving mouse primitive endoderm formation, pancreatic development, and haematopoiesis. For both simulated and experimental data, we find that locaTE provides a powerful, efficient and scalable network inference method that allows us to distil cell-specific networks from single cell data.
Competing Interest Statement
The authors have declared no competing interest.
Footnotes
Restructuring of manuscript and additional results prior to submission