Evaluating CNV Detection Software Performance
Author Information
Author(s): Zhang Dandan, Qian Yudong, Akula Nirmala, Alliey-Rodriguez Ney, Tang Jinsong, Gershon Elliot S., Liu Chunyu
Primary Institution: Department of Psychiatry and Behavioral Neuroscience, The University of Chicago
Hypothesis
How accurately can different software detect copy number variants (CNVs) from genome-wide association study (GWAS) data?
Conclusion
Birdsuite outperformed other software in detecting rare CNVs, but all programs showed poor accuracy for common CNVs.
Supporting Evidence
- Birdsuite recovered the highest percentages of known HapMap CNVs.
- The recovery rate increased with decreased CNV frequency.
- Birdsuite's call was 98.8% consistent with qPCR quantification in one CNV region.
Takeaway
This study looked at different computer programs to see which one is best at finding changes in DNA that might be linked to diseases. Birdsuite was the best at finding rare changes, but all programs struggled with common ones.
Methodology
The study compared four CNV detection software programs using data from HapMap samples and the Bipolar Genome Study, assessing recovery rates and validation through qPCR.
Potential Biases
The software-specific calls tended not to be validated, indicating potential bias in program-specific results.
Limitations
The study had a limited number of CNVs tested by qPCR, particularly for common CNVs.
Participant Demographics
1001 bipolar cases and 1033 controls, all of European ancestry.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website