DIscBIO

DIscBIO performs biomarker discovery from single-cell transcriptomics by integrating scRNA-seq analyses including clustering, differential expression, decision tree-based biomarker selection and gene enrichment within a network context.

Key Features:

• Multi-algorithmic integration: Combines multiple scRNA-seq packages to execute integrated single-cell analyses.
• Clustering: Performs clustering of single-cell sequencing read counts to identify cellular sub-populations.
• Differential expression analysis: Conducts differential expression analysis between clusters or experimental conditions.
• Decision tree-based biomarker discovery: Applies decision tree algorithms for selection of candidate biomarkers.
• Gene enrichment analysis: Performs gene set enrichment analysis to identify enriched pathways or functions.
• Network context analysis: Integrates enrichment and biomarker results within a network context for biological interpretation.

Scientific Applications:

• Circulating tumor cells (breast cancer): Demonstrated on circulating tumor cells from breast cancer patients to identify molecular signatures of cellular sub-populations.
• Cell cycle regulation (myxoid liposarcoma): Applied to a cell cycle regulation dataset in myxoid liposarcoma to characterize molecular signatures.
• Oncology and single-cell transcriptomics: Used to identify molecular signatures characterizing cellular sub-populations in oncology and other single-cell transcriptomics studies.

Methodology:

Processes single-cell sequencing read counts through clustering and differential expression analysis, followed by decision tree-based biomarker discovery and gene enrichment analysis.