Bayesian Estimation of Genomic Copy Number Using SNP Genotyping Arrays
Author Information
Author(s): Guo Beibei, Villagran Alejandro, Vannucci Marina, Wang Jian, Davis Caleb, Man Tsz-Kwong, Lau Ching, Guerra Rudy
Primary Institution: Rice University
Hypothesis
Can a Bayesian spatial normal mixture model effectively estimate genomic copy numbers from SNP genotyping arrays?
Conclusion
The Bayesian mixture model performs comparably to the CNAG algorithm and sometimes outperforms it, especially in accounting for normal cell contamination in tumor samples.
Supporting Evidence
- The Bayesian model can adjust for normal cell contamination, which is a common issue in tumor samples.
- Simulation studies showed that the model can accurately detect both long and short segments of copy number variations.
- The method provides smoother results compared to traditional hidden Markov models.
Takeaway
This study created a new method to find changes in DNA that can help understand cancer better, and it works well even when normal cells mix with cancer cells.
Methodology
The study used a Bayesian spatial normal mixture model and a reversible jump Markov chain Monte Carlo algorithm to analyze SNP data.
Potential Biases
Potential bias due to normal cell contamination affecting copy number inference.
Limitations
The method may not perform well with very small segments of copy number variations.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website