CKAT

CKAT performs composite kernel association testing to jointly assess genetic main effects and gene-treatment interaction effects for sets of single-nucleotide polymorphisms (SNPs) in pharmacogenetic analyses of randomized clinical trials.

Key Features:

• Joint Testing Capability: Enables simultaneous evaluation of genetic main effects and SNP-treatment interaction effects for SNP sets using a kernel machine framework.
• Efficiency in Whole Exome/Genome Analysis: Designed to handle whole-exome and whole-genome association analyses with attention to computational efficiency and control of type I error in small clinical-trial sample sizes.
• Analytic P-value Calculation: Implements an analytic procedure for calculating P-values that avoids permutation or perturbation tests.
• Type I Error Control and Power Performance: Demonstrated control of type I error rates and improved power performance across simulation studies and practical applications.
• Application Example: Applied to gene-level association testing for reduction of Clostridium difficile infection recurrence in patients treated with bezlotoxumab.

Scientific Applications:

• Pharmacogenetic association testing: Assess cumulative effects of multiple SNPs on drug response to identify genetic markers that can inform personalized treatment strategies.
• Clinical trial gene-treatment interaction analysis: Test gene-treatment interactions and genetic main effects within randomized clinical trials, including gene-level analyses in therapeutic studies such as bezlotoxumab for Clostridium difficile.

Methodology:

Uses kernel machine-based models and an analytic P-value calculation to jointly test genetic main effects and gene-treatment interactions for SNP sets.