npSeq

npSeq performs significance analysis of count-based sequencing data using a non-parametric approach to identify differentially expressed features in RNA-Seq and comparative genomic experiments.

Key Features:

• Non-Parametric Approach: Uses a non-parametric methodology instead of Poisson or negative binomial models to analyze RNA-Seq count data that may not meet parametric assumptions.
• Resampling Technique: Incorporates resampling techniques to account for differences in sequencing depths across experiments and enable more reliable comparisons between datasets with varying total read counts.
• Symmetric Cutoffs: Employs symmetric cutoffs for significance determination, contrasting with asymmetric cutoffs used by tools such as SAM 4.0, to identify both up-regulated and down-regulated genes.

Scientific Applications:

• Quantitative Outcomes: Analyzes associations between gene expression and continuous variables in genomic studies.
• Survival Analysis: Investigates gene expression patterns related to survival times or event-based outcomes.
• Classification Tasks: Distinguishes features associated with two-class or multiple-class outcomes, such as disease versus healthy states.

Methodology:

npSeq applies a non-parametric comparative analysis using resampling to adjust for sequencing depth and employs symmetric cutoffs to detect up- and down-regulated features.