Back

Addressing multiple facets of ligand-receptor network inference including single-cell proteomics

Villemin, J.-P.; Giroux, P.; Maillard, M.; Colombo, P.-E.; Larbouret, C.; Colinge, J.

2025-10-06 bioinformatics
10.1101/2025.10.05.680519 bioRxiv
Show abstract

Distinct ligand-receptor interaction (LRI) inference tools often produce markedly different results, and their performance can vary considerably across datasets. Indeed, performance is influenced by differences in experimental designs and dataset-specific features, making it difficult to establish a universal LRI tool. To address this challenge, we expanded our SingleCellSignalR Bioconductor package to provide an integrated framework that incorporates alternative scoring strategies and adjustable analytical depth. We motivate this choice through the analysis of two single-cell transcriptomics datasets that exemplify contrasting experimental designs. Leveraging the new framework flexibility, we present a detailed analysis of paired single-cell proteomics and transcriptomics data, providing, to our knowledge, the first direct comparison of LRI inference across these complementary modalities at single-cell resolution. Finally, we demonstrate how the same framework seamlessly accommodates additional underexplored data types from the LRI perspective, including patient-derived mouse xenografts and bulk RNA sequencing of upstream-sorted cell populations.

Published in Nucleic Acids Research (predicted rank #5) · training set

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.