Analyzing Genetic Markers for Rheumatoid Arthritis
Author Information
Author(s): Beate Glaser, Ivan Nikolov, Daniel Chubb, Marian L Hamshere, Ricardo Segurado, Valentina Moskvina, Peter Holmans
Primary Institution: Cardiff University
Hypothesis
How do single locus and pairwise effects of genetic markers relate to rheumatoid arthritis?
Conclusion
The study found a strong correspondence between random forest importance measures and logistic regression interaction effects, suggesting that random forests could be useful for pre-screening in larger genetic studies.
Supporting Evidence
- The study analyzed 20 markers from the Genetic Analysis Workshop 15 dataset.
- Random forests showed strong correlations with logistic regression results.
- The most consistent pairwise effect was found between markers rs237025 and rs577001.
Takeaway
Researchers looked at genetic markers to see how they might work together to cause rheumatoid arthritis. They found that some markers are really important and can help in future studies.
Methodology
The study used logistic regression and random forests to analyze 20 genetic markers in a dataset of rheumatoid arthritis patients and controls.
Potential Biases
Data imputation methods may introduce bias, especially if the underlying genotype distribution is not carefully investigated.
Limitations
The study's findings may not be generalizable due to the small sample size and the potential for false positives in the random forest analysis.
Participant Demographics
Participants were independent individuals of Caucasian origin, including 463 cases and 855 controls.
Statistical Information
P-Value
p = 2 × 10-8 for rs2476601
Want to read the original?
Access the complete publication on the publisher's website