Identifying Hypermethylated CpG Islands with a Quantile Regression Model
Author Information
Author(s): Sun Shuying, Chen Zhengyi, Yan Pearlly S, Huang Yi-Wen, Huang Tim HM, Lin Shili
Primary Institution: Case Comprehensive Cancer Center, Case Western Reserve University
Hypothesis
Which quantile level should be used to identify hypermethylated CGIs and their associated genes?
Conclusion
The quantile levels ranging from 80% to 90% are better at identifying known methylated and unmethylated genes.
Supporting Evidence
- The quantile regression model was applied to methylation microarray data from breast cancer cell lines.
- Results indicated that quantile levels between 80% and 90% are optimal for identifying hypermethylated regions.
- Statistical measurements showed consistent patterns across breast and ovarian cancer data.
Takeaway
The researchers found that using a special math model can help identify areas in DNA that are overly methylated, which is important for understanding cancer.
Methodology
The study used two DMH microarray datasets from 40 breast cancer cell lines and 26 ovarian cancer patients, applying a quantile regression model to identify hypermethylated CGIs.
Potential Biases
Potential confounding factors affecting observed methylation signals were noted, including probe effects and noise.
Limitations
The study's results may vary due to differences in sample types and sizes between breast cancer cell lines and ovarian cancer patients.
Participant Demographics
40 breast cancer cell lines and 26 ovarian cancer patients.
Statistical Information
P-Value
0.05
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website