reconsi

reconsi improves simultaneous inference under dependence of tests by estimating a collapsed null distribution through resampling.

Key Features:

• Dependence-aware inference: Accounts for dependence between tests to increase power while reducing the variability of the false discovery proportion.
• Collapsed null estimation: Estimates a collapsed null distribution through resampling to improve simultaneous inference.

Scientific Applications:

• Flow cytometry and microbiome integration: Enables simultaneous inference when combining flow cytometry measurements with microbiome sequence counts.
• Genomics applications: Addresses dependence structures common in genomics data where tests are not independent.

Methodology:

Estimates a collapsed null distribution through resampling to account for dependence between tests.