# MDIC3: Matrix decomposition to infer cell-cell communication

## TL;DR

MDIC3 is a new tool that can infer how cells communicate with each other using single-cell data, without needing prior knowledge of ligand-receptor pairs.

## Contribution

MDIC3 is the first unsupervised method for cell-cell communication inference that works across species and does not rely on ligand-receptor databases.

## Key findings

- MDIC3 outperforms existing methods in inferring cell-cell communication in both human and mouse data.
- MDIC3 successfully infers cell-cell communication in zebrafish and human skin samples.
- MDIC3 separates gene and cell information from expression data without using prior ligand-receptor knowledge.

## Abstract

Crosstalk among cells is vital for maintaining the biological function and intactness of systems. Most existing methods for investigating cell-cell communications are based on ligand-receptor (L-R) expression, and they focus on the study between two cells. Thus, the final communication inference results are particularly sensitive to the completeness and accuracy of the prior biological knowledge. Because existing L-R research focuses mainly on humans, most existing methods can only examine cell-cell communication for humans. As far as we know, there is currently no effective method to overcome this species limitation. Here, we propose MDIC3 (matrix decomposition to infer cell-cell communication), an unsupervised tool to investigate cell-cell communication in any species, and the results are not limited by specific L-R pairs or signaling pathways. By comparing it with existing methods for the inference of cell-cell communication, MDIC3 obtained better performance in both humans and mice.

•MDIC3 can be used to infer cell-cell communication without prior L-R information•MDIC3 relies only on single-cell sequencing data and is species independent•MDIC3 separates information about genes and cells from expression profiles•MDIC3 contributes to improved cell-cell communication inference resolution

MDIC3 can be used to infer cell-cell communication without prior L-R information

MDIC3 relies only on single-cell sequencing data and is species independent

MDIC3 separates information about genes and cells from expression profiles

MDIC3 contributes to improved cell-cell communication inference resolution

Cell-cell communication (CCC) can reveal diverse aspects of life processes and cellular function. CCC involves information exchange among cells via ligand-receptor (L-R) pairs. Existing CCC studies rely on gathering L-R pairs to construct databases and analyze L-R gene co-expression. Building comprehensive L-R databases for different species, however, is challenging, especially considering the sensitivity of CCC inference to database completeness and accuracy. Methods of CCC that do not rely on L-R lists will have opportunities to discover more comprehensive CCC across different species. This holds great practical significance for exploring the functional interactions among cells in different biological states, understanding the biological reasons behind the therapeutic effects of various drugs on diseases for drug development, and elucidating CCC between different species (such as understanding the CCC among hosts and microorganisms to elucidate infection mechanisms).

The authors present MDIC3, an unsupervised method for inferring cell-cell communication without prior ligand-receptor information. MDIC3 separates information about genes and cells directly from expression matrices derived from single-cell sequencing data. The authors demonstrate the use of the method in mouse and zebrafish embryonic development and in skin samples from patients with atopic dermatitis.