EmpiricalBrownsMethod

EmpiricalBrownsMethod combines dependent P-values from correlated high-throughput biological experiments by applying an empirical adaptation of Brown's Method, extending Fisher's Method.

Key Features:

• Handling Dependent P-values: Combines dependent P-values and accounts for the correlation structures typical of high-throughput biological data.
• Empirical Adaptation: Extends Brown's Method with empirical adjustments to better model dependencies among tests.
• Performance Superiority: Demonstrates improved performance over Fisher's Method and alternative approaches on noisy simulated datasets and real-world gene expression data such as The Cancer Genome Atlas.
• Implementations: Implemented in Python, R, and MATLAB.
• Bioconductor Integration: R implementation distributed as a Bioconductor package (EmpiricalBrownsMethod).

Scientific Applications:

• Genomic Studies: Analyzing gene expression data from The Cancer Genome Atlas.
• Simulated Data Analysis: Evaluating robustness and reliability of statistical combination methods on noisy simulated datasets.

Methodology:

Uses an empirical adaptation of Brown's Method (an extension of Fisher's Method) with empirical adjustments and validation on noisy simulated and real-world gene expression datasets such as The Cancer Genome Atlas.