Copy number variation accuracy in genome-wide association studies

Area of Science:

• Genetics
• Bioinformatics
• Genomics

Background:

• Copy number variations (CNVs) are significant genetic alterations linked to various diseases.
• Identifying CNVs using genome-wide association studies and microarrays is crucial but faces reliability challenges.
• Existing computational tools for CNV detection often produce questionable results.

Purpose of the Study:

• To evaluate the performance of different CNV detection algorithms.
• To develop a statistical method for estimating the sensitivity and positive predictive values of CNV calls.
• To enhance the reliability of CNV identification in genome-wide data.

Main Methods:

• Comparative analysis of CNVPartition, PennCNV, and QuantiSNP software.
• Development and application of a statistical method for assessing CNV call accuracy.
• Validation using 96 duplicate samples within the dataset.

Main Results:

• Positive predictive rates for CNVs increase with probe number and physical size (>= 500 kb, >= 100 probes).
• Utilizing CNVs identified by multiple programs significantly improves reproducibility and positive predictive rates.
• The performance of CNV detection tools varies, necessitating careful selection and validation.

Conclusions:

• The developed methods enable more reliable identification of CNVs from genome-wide data.
• Researchers can improve CNV detection accuracy by considering CNV size and consensus calls from multiple algorithms.
• This approach aids in reducing false positives and increasing confidence in CNV findings for disease association studies.