Related Experiment Video
Updated: Aug 12, 2026
11:52
Targeted RNA Sequencing Assay to Characterize Gene Expression and Genomic Alterations
Published on: August 4, 2016
A topology-based framework for robust cancer-associated gene signature identification from scRNA-seq data
Sudarshan Gogoi1, Subhajit Bandyopadhyay2, Soumen Bera3
1Department of Mathematics, Sikkim University, Gangtok, Sikkim, India.
Computational Biology and Chemistry
|August 10, 2026
Summary
This study introduces a novel topology-guided framework for cancer transcriptomics. It uncovers hidden gene signatures in single-cell RNA sequencing data, improving cancer biomarker discovery and precision oncology.
Area of Science:
- Computational biology
- Cancer research
- Genomics
Background:
- Conventional gene selection in cancer transcriptomics struggles with high-dimensional single-cell RNA sequencing (scRNA-seq) data, failing to capture geometric structures and often obscured by noise and heterogeneity.
- Existing methods focus on statistical variance, neglecting the underlying topological features crucial for identifying cancer-specific signals.
Purpose of the Study:
- To develop and validate a topology-guided framework using persistent homology for extracting structurally invariant, cancer-associated gene signatures from scRNA-seq data.
- To move beyond traditional gene-level statistics towards shape-aware biological discovery in cancer transcriptomics.
- To enhance biomarker discovery for precision oncology and next-generation cancer diagnostics.
Main Methods:
- Integration of highly variable feature selection and dimensionality reduction with Vietoris-Rips filtration-based gene correlation topology.
- A dual-stage stability-driven classification strategy to identify reproducible cancer-specific topological patterns.
- Rigorous validation of topologically significant genes using differential expression analysis, ROC/AUC, KEGG pathway enrichment, PPI networks, and literature evidence.
Main Results:
- The topology-guided framework demonstrated superior discriminative power and biological relevance compared to conventional HVF+PCA methods.
- Identified cancer-type-specific topological patterns, revealing a mitotic regulatory module in breast cancer and ECM remodeling/TME mechanisms in colorectal cancer.
- Discovered novel candidate biomarkers with topological and statistical significance, not present in standard pathway databases.
Conclusions:
- Persistent homology offers a transformative, structure-aware paradigm for transcriptomic biomarker discovery in cancer.
- The framework provides a principled foundation for precision oncology, yielding more focused pathway enrichment and functionally coherent gene sets.
- This approach significantly advances the ability to decode complex scRNA-seq data for improved cancer diagnostics and treatment strategies.
