CNVassoc: Association analysis of CNV data using R.

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Copy number variants (CNVs) are significant contributors to complex disease genetic risk.
• Accurate analysis of CNV-disease associations requires accounting for uncertainty in copy-number calls.
• Existing statistical tools may yield biased results if CNV call uncertainty is ignored.

Purpose of the Study:

• To introduce CNVassoc, an R package designed for robust association analysis of common CNVs.
• To provide a flexible tool that accommodates various response variables, study designs, and inheritance models.
• To enable the incorporation of uncertainty in CNV genotype calling within association analyses.

Main Methods:

• Developed CNVassoc, an R package for population-based CNV association studies.
• Included functions for CNV genotype calling and acceptance of data from other algorithms (e.g., CANARY, CGHcall).
• Implemented methods to test associations with diverse response variables (e.g., case-control status, censored data, counts) and adjust for covariates.

Main Results:

• CNVassoc handles various CNV types from different platforms (MLPA, aCGH).
• Demonstrated the package's ability to incorporate uncertainty in association analysis using real data.
• Simulation studies showed CNVassoc outperforms CNVtools in computation time and convergence rates.

Conclusions:

• CNVassoc offers superior modeling flexibility, statistical power, and convergence rates compared to existing packages.
• The package simplifies covariate adjustment and has lower requirements for sample size and signal quality.
• CNVassoc is recommended for routine use in copy number variant association studies.