BlackSheep

BlackSheep performs differential extreme-value analysis on genome-wide molecular assay data to identify and characterize extreme measurements in high-resolution assays such as shotgun proteomics and RNA-seq, with emphasis on small or underrepresented sample subgroups.

Key Features:

• Differential extreme-value analysis: Performs differential analysis that targets extreme measurements rather than central tendencies.
• Nonparametric approach: Employs a nonparametric method for description and differential analysis, reducing reliance on parametric assumptions.
• Genome-wide compatibility: Operates on genome-wide data derived from high-resolution molecular characterization assays.
• Assay support: Specifically applicable to shotgun proteomics and RNA-seq datasets.
• Sensitivity to small subgroups: Designed to detect signals in small or rare subgroups within larger cohorts, such as rare disease subtypes.
• Complementary to differential expression methods: Complements existing differential expression analyses by prioritizing extreme-value signals that may be missed by standard approaches.
• Robustness to distributional heterogeneity: Robust against the varied distributions commonly observed in high-resolution molecular data.

Scientific Applications:

• Rare subgroup biomarker discovery: Identifying biomarkers or signals enriched in rare disease subtypes or underrepresented groups within cancer cohorts.
• Proteomics outlier detection: Detecting extreme protein expression events in shotgun proteomics datasets.
• Transcriptomics outlier detection: Detecting extreme expression outliers in RNA-seq data that may indicate biologically significant events.

Methodology:

Uses a nonparametric approach to describe and perform differential analysis of genome-wide data by focusing on extreme measurements.