multiple-testing-correction

OmicsQ is a Shiny-based web application designed for preprocessing and quantitative analysis of Omics data. It streamlines the preparation of data for subsequent analysis using specialized tools such as PolySTest for statistical testing and VSClust for clustering analysis.

Python package for mechanistic interpretability providing statistical audits of activation patching for TransformerLens with False Discovery Rate (FDR) control, Benjamini-Hochberg correction, and sign-flip p-values.

Multiple Hypotheses Testing for Discrete Data

This is the second integrated (mid-term) project, in which I had to analyze sales funnel and test multiple hypotheses based on the results of A/A/B test.

A package for aggregating p-values.

Evaluation of Simultaneous Inference Methods for the Human Genome.

780 pre-registered multi-asset backtests, corrected for multiple testing. 0 survive at 5%.

Extensive collection of resources on multiple hypothesis testing.

Mémoire détaillé sur les tests multiples en biostatistique.