signifinder

signifinder computes and explores a compendium of tumor transcriptional signatures from public literature, returning single-sample, single-cell, or single-spot scores based on gene expression values across bulk, single-cell, and spatial transcriptomics data.

Key Features:

• Signature Collection: The package currently collects more than 70 distinct signatures relating to multiple tumors and multiple cancer processes.
• Multi-Platform Compatibility: signifinder computes signatures across bulk RNA sequencing, single-cell RNA sequencing, and spatial transcriptomics data.
• Single-Sample Scoring: The package returns single-sample, single-cell, or single-spot scores for each computed signature.
• Tumor Characterization: Users can assess hallmark processes, therapy responses, and tumor microenvironment peculiarities using curated publicly available signatures.

Scientific Applications:

• Cancer Mechanism Exploration: signifinder enables investigation of cancer mechanisms through transcriptional signature analysis across multiple data types.
• Subtype Definition: The package supports cancer subtype classification using established gene expression signatures.
• Prognosis Prediction: signifinder facilitates prognostic assessment through computation of published prognostic signatures.
• Therapy Efficacy Assessment: Users can evaluate therapy response signatures to predict treatment outcomes.
• Tumor Heterogeneity Analysis: The package characterizes tumor mass heterogeneity at cellular resolution in single-cell and spatial transcriptomics data.

Methodology:

signifinder computes signature scores based on gene expression values from user-provided data, applying curated signature gene sets from public literature to return single-sample, single-cell, or single-spot scores.