Celcomen: spatial causal disentanglement for single-cell and tissue perturbation modeling
This provides a method to model disease and therapy-induced changes in spatially resolved tissues, offering insights relevant to human health, though it appears incremental as it builds on existing causality and GNN frameworks.
The paper tackled the problem of disentangling intra- and inter-cellular gene regulation in spatial transcriptomics and single-cell data using a generative graph neural network, resulting in validated capabilities for counterfactual predictions in clinical samples like glioblastoma and fetal spleen.
Celcomen leverages a mathematical causality framework to disentangle intra- and inter- cellular gene regulation programs in spatial transcriptomics and single-cell data through a generative graph neural network. It can learn gene-gene interactions, as well as generate post-perturbation counterfactual spatial transcriptomics, thereby offering access to experimentally inaccessible samples. We validated its disentanglement, identifiability, and counterfactual prediction capabilities through simulations and in clinically relevant human glioblastoma, human fetal spleen, and mouse lung cancer samples. Celcomen provides the means to model disease and therapy induced changes allowing for new insights into single-cell spatially resolved tissue responses relevant to human health.