Predicting Psoriasis Using SNP Data
Author Information
Author(s): Fang Shenying, Fang Xiangzhong, Xiong Momiao
Primary Institution: The University of Texas M. D. Anderson Cancer Center
Hypothesis
Can single nucleotide polymorphisms (SNPs) be used to predict susceptibility to psoriasis?
Conclusion
A small set of SNPs can predict disease status with an average accuracy of 68%, indicating the potential for SNP data in psoriasis prediction.
Supporting Evidence
- The best test harmonic mean of sensitivity and specificity for predicting psoriasis by sIB was 0.674.
- The study compared the performance of sIB with LDA, showing sIB performed better.
- A total of 2,798 samples and 451,724 SNPs were analyzed.
Takeaway
Scientists found that certain tiny changes in our DNA can help predict if someone will get psoriasis, which is a skin disease.
Methodology
The study used a dataset of SNPs to develop a classifier for predicting psoriasis, comparing the performance of the sequential information bottleneck (sIB) method with linear discriminant analysis (LDA).
Potential Biases
Potential overfitting due to the selection of SNPs based on performance metrics.
Limitations
The study's results may not generalize beyond the specific populations studied, and the identified SNPs may not be causally related to psoriasis.
Participant Demographics
The study included 1,627 subjects in the general research use group and 1,171 subjects in the autoimmune disease only group, with cases diagnosed by dermatologists.
Statistical Information
P-Value
0
Confidence Interval
95% CI: 0.650-0.698
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website