CODEX2: full-spectrum copy number variation detection by high-throughput DNA sequencing.

Area of Science:

• Genomics
• Bioinformatics
• Statistical Genetics

Background:

• High-throughput DNA sequencing offers high resolution for detecting genome-wide copy number variations (CNVs).
• Existing sequencing methods are prone to biases and artifacts, leading to reduced sensitivity and false discoveries.
• Array-based methods for CNV detection have lower resolution compared to sequencing.

Purpose of the Study:

• To introduce CODEX2, a statistical framework for comprehensive CNV profiling using DNA sequencing data.
• To develop a method sensitive to both common and rare CNVs across various population frequencies.
• To evaluate CODEX2's performance on whole-exome and targeted sequencing data, particularly in the presence of biases.

Main Methods:

• Development of CODEX2, a statistical framework for CNV detection.
• Application and evaluation of CODEX2 on whole-exome sequencing (WES) and targeted sequencing datasets.
• Comparison of CODEX2 with existing CNV detection methods.

Main Results:

• CODEX2 demonstrates high sensitivity for detecting CNVs with both common and rare population frequencies.
• The framework effectively handles study designs with or without negative control samples.
• CODEX2 significantly improves sensitivity for detecting common CNVs compared to existing methods.
• Performance was validated on whole-exome and targeted sequencing data, where biases are most prominent.

Conclusions:

• CODEX2 provides a robust statistical framework for accurate and sensitive CNV profiling from DNA sequencing.
• The method overcomes limitations of existing approaches, particularly for common CNVs.
• CODEX2 is applicable to diverse sequencing study designs and data types, enhancing genomic variation analysis.