A Comparison of Cell-Cell Interaction Prediction Tools Based on scRNA-seq Data.

Area of Science:

• Computational biology
• Immunoinformatics
• Genomics

Background:

• Understanding cell-cell interactions (CCIs) is crucial for disease progression insights.
• Single-cell RNA sequencing (scRNA-seq) enables detailed analysis of cellular communication.

Purpose of the Study:

• To benchmark and evaluate the performance of various computational tools for predicting CCIs from scRNA-seq data.
• To identify the most effective tools for CCI prediction in research settings.

Main Methods:

• Comparative analysis of multiple CCI prediction tools: CCInx, CellChat, CellPhoneDB, iTALK, NATMI, scMLnet, SingleCellSignalR, and an ensemble method.
• Validation against a manually curated gold standard dataset for idiopathic pulmonary fibrosis (IPF).
• Evaluation metrics included prediction accuracy and computational processing time.

Main Results:

• CellPhoneDB and NATMI demonstrated superior performance in predicting CCIs, defined as source-target-ligand-receptor tetrads.
• Performance varied among tools, with significant differences in processing times.
• Specific tool recommendations were provided based on research project needs.

Conclusions:

• CellPhoneDB and NATMI are recommended as top-performing tools for CCI prediction from scRNA-seq data.
• The study provides guidance for researchers selecting appropriate CCI prediction tools.
• Future directions for improving CCI prediction methodologies were discussed.