scKGOT: Intercellular Signaling Inference with Knowledge Graph Optimal Transport for Single-cell Transcriptomics
Keywords: Knowledge Graph, Optimal Transport, Cell-cell Communication
TL;DR: scKGOT uses Knowledge Graph Optimal Transport to model intercellular signaling with guidance from the Ligand-Receptor-Pathway Knowledge Graph, improving pathway inference, precision, and interpretability while revealing cellular communication.
Abstract: Single-cell transcriptomics provides detailed genetic insights into cellular heterogeneity within intact organs and the intercellular signaling that underpins tissue homeostasis, development, and disease. To improve the inference of intercellular signaling and pathway activity, we introduce scKGOT, a novel method that employs the Knowledge Graph Optimal Transport (KGOT) algorithm to model and quantify ligand-receptor-signaling networks between sender and receiver cells. scKGOT defines sender and receiver spaces using pairwise distance matrices from gene expression profiles and leverages prior knowledge from the Ligand-Receptor-Pathway Knowledge Graph (LRP-KG) as initial guidance for transport optimization, allowing for dynamic adaptation based on gene expression data. Through comprehensive benchmarking on public single-cell transcriptomic datasets, scKGOT consistently outperforms existing inference methods in terms of precision and interpretability. Furthermore, we demonstrate its practical applicability across multiple case studies, uncovering complex pathway interactions and revealing insights into cellular heterogeneity in diverse biological contexts. By incorporating scKGOT, we provide a robust and generalizable approach for pathway inference in single-cell analyses, advancing the understanding of intercellular communication mechanisms and offering valuable insights into biological processes at the cellular level.
Primary Area: applications to physical sciences (physics, chemistry, biology, etc.)
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 14080
Loading