Clipper

Clipper controls false discovery rate (FDR) in high-throughput biological comparisons between two conditions without using p-values or requiring specific distributional assumptions.

Key Features:

• P-Value Independence: Operates independently of p-values, avoiding reliance on p-value computation for significance assessment.
• Distributional Robustness: Provides valid FDR control when p-value computation is challenging due to distributional assumptions or limited replicates.
• General Statistical Framework: Adapts to multiple high-throughput data types, including genes, genomic regions, and proteins.
• Enhanced Performance: Has been shown to outperform existing methods across diverse high-throughput applications in controlling FDR.

Scientific Applications:

• Differential Feature Identification: Identifying features with differential values between two conditions in large-scale datasets.
• Genomics: Identifying differentially expressed genes or genomic regions under varying experimental conditions.
• Proteomics: Detecting proteins that exhibit significant changes in abundance between two states.

Methodology:

A general statistical framework that controls FDR without relying on p-values by leveraging alternative p-value-free criteria to identify significant features between two conditions.