fdrci

fdrci provides permutation-based selection and adjustment of false discovery rate (FDR) confidence intervals to enable post hoc identification of alternative discovery thresholds and detection of likely true positives in large-scale hypothesis testing.

Key Features:

• Permutation-based FDR estimator: Employs a permutation-based FDR estimator to assess significance across multiple hypothesis tests.
• FDR confidence interval selection and adjustment: Selects and adjusts FDR confidence intervals to evaluate the robustness of discovered signals.
• Interval exclusion criterion: Identifies FDR confidence intervals that do not encompass the value one as indicators of likely true positives.
• Post hoc discovery threshold determination: Determines multiple potential discovery thresholds post hoc, including thresholds beyond the conventional 0.05.

Scientific Applications:

• Transcriptome-wide association study (MAVERICC clinical trial): Applied to a TWAS in the MAVERICC trial to identify genes whose predicted expression levels were associated with progression-free or overall survival in patients with metastatic colorectal cancer.
• Large-scale hypothesis testing in genomics and clinical research: Facilitates exploration of weak but potentially meaningful effects by evaluating alternative discovery thresholds in large-scale genomic and clinical studies.

Methodology:

Uses a permutation-based FDR estimator with FDR confidence interval selection and adjustment, identifies intervals that do not include the value one, and determines multiple potential discovery thresholds post hoc.