Related Experiment Video
Updated: Jun 25, 2025
10:12
Droplet Barcoding-Based Single Cell Transcriptomics of Adult Mammalian Tissues
Published on: January 10, 2019
18.5K
Clustering and visualization of single-cell RNA-seq data using path metrics
Andriana Manousidaki1, Anna Little2, Yuying Xie1,3
1Department of Statistics and Probability, Michigan State University, East Lansing, Michigan, United States of America.
Plos Computational Biology
|May 29, 2024
Summary
Single-cell Path Metrics Profiling (scPMP) is a new framework that accurately analyzes tissue and cancer cell data. It preserves both local and global data structures, outperforming existing single-cell RNA sequencing clustering methods.
Area of Science:
- Computational Biology
- Genomics
- Bioinformatics
Background:
- Single-cell technologies offer high-resolution insights into tissue and cancer composition.
- Existing dimension reduction and clustering tools struggle to preserve both local and global data structures.
Purpose of the Study:
- To develop a novel analysis framework, Single-Cell Path Metrics Profiling (scPMP), for single-cell data.
- To address limitations in preserving local cluster structure and global data geometry.
Main Methods:
- Developed scPMP using power-weighted path metrics for data-driven distance measurement.
- Employed multidimensional scaling to create low-dimensional embeddings.
- Path metrics are density-sensitive and respect underlying data geometry, unlike Euclidean distance.
Main Results:
- scPMP effectively preserves both global data geometry and cluster structure.
- Evaluated clustering quality and geometric fidelity.
- Demonstrated superior performance compared to current scRNAseq clustering algorithms across various datasets.
Conclusions:
- scPMP offers a robust framework for single-cell data analysis.
- The method enhances the preservation of data topology, leading to improved clustering.
- scPMP represents a significant advancement for analyzing complex single-cell datasets.
