mastR

mastR automates the identification of tissue-specific gene signatures from labeled bulk RNA-seq data by integrating differential expression analysis with background noise correction.

Key Features:

• Automated marker identification: Generates refined lists of signature genes from multiple group comparisons without requiring manual curation of results or statistical thresholds.
• Differential expression integration: Leverages edgeR and limma statistical pipelines to perform pairwise comparisons between groups.
• Rank-product-based aggregation: Aggregates pairwise comparison results using a rank-product-based permutation test.
• Background noise minimization: Implements a signal-to-noise ratio approach to account for tissue background expression and minimize background signals.
• Tissue-specificity filtering: Explicitly considers tissue-specific background noise, which is often ignored by other marker generation tools.

Scientific Applications:

• Cancer research: Identifies group markers for tumor classification and biomarker discovery in cancer studies.
• Developmental biology: Identifies group markers for studying developmental processes and cell type differentiation.
• Immune cell profiling: Derives cell type-specific signatures, as demonstrated with NK cell signature identification and validation against published curated signatures.
• Single-cell RNA-seq analysis: Demonstrated utility on simulated scRNA-seq data and comparison with Seurat for marker selection performance.

Methodology:

mastR uses edgeR and limma to perform pairwise comparisons between groups, aggregates results using a rank-product-based permutation test, and applies a signal-to-noise ratio approach to minimize background signals.