Related Experiment Video
Updated: Aug 25, 2025
Spatial Profiling of Protein and RNA Expression in Tissue: An Approach to Fine-Tune Virtual Microdissection
Published on: July 6, 2022
Evaluation of cell-cell interaction methods by integrating single-cell RNA sequencing data with spatial information
Zhaoyang Liu1,2, Dongqing Sun1,2, Chenfei Wang3,4
1Key Laboratory of Spine and Spinal Cord Injury Repair and Regeneration of Ministry of Education, Department of Orthopedics, Tongji Hospital, School of Life Science and Technology, Tongji University, Shanghai, 200092, China.
Evaluating cell-cell interaction tools using spatial transcriptomics data is crucial for understanding biological processes. This study benchmarks 16 methods, finding statistical approaches superior and recommending CellChat, CellPhoneDB, NicheNet, and ICELLNET for accuracy.
Area of Science:
- Computational Biology
- Genomics
- Cell Biology
Background:
- Cell-cell interactions are vital for biological processes, with computational methods using single-cell RNA sequencing (scRNA-seq) emerging for their characterization.
- Evaluating these computational tools is challenging due to the lack of ground truth data.
- Spatial transcriptomics (ST) provides cell positional information, offering a potential basis for interaction assessment.
Purpose of the Study:
- To develop and apply a novel evaluation framework for cell-cell interaction prediction tools.
- To benchmark 16 existing computational methods using integrated scRNA-seq and ST data.
- To assess the performance and consistency of various cell-cell interaction tools.
Main Methods:
- Integrated scRNA-seq and ST data from 15 simulated and 5 real datasets.
- Classified cell-cell interactions into short-range and long-range based on spatial distance distributions.
- Defined a distance enrichment score for evaluating tool performance.
Main Results:
- Benchmarked 16 cell-cell interaction tools, revealing highly dynamic interaction predictions.
- Statistical-based methods outperformed network-based and ST-based methods.
- Identified CellChat, CellPhoneDB, NicheNet, and ICELLNET as top-performing tools based on spatial consistency and scalability.
Conclusions:
- The study provides a comprehensive evaluation of scRNA-seq-based cell-cell interaction tools.
- Recommends using multiple methods for robust interaction identification and highlights top-performing tools.
- A benchmark workflow is available on GitHub for community use.
