Selecting Important SNPs for Population Classification
Author Information
Author(s): Zhou Nina, Wang Lipo
Primary Institution: Nanyang Technology University
Hypothesis
Can we effectively select a small number of informative SNPs from the HapMap genotype data to classify populations?
Conclusion
The study demonstrates that using a modified t-test and F-statistics can effectively identify a small number of SNPs that accurately classify populations.
Supporting Evidence
- The modified t-test and F-statistics methods were effective in ranking SNPs.
- Using an average of 64 SNPs, the study achieved good classification of three populations.
- The modified t-test provided higher classification accuracy compared to F-statistics.
Takeaway
Scientists found a way to pick just a few important genetic markers to tell different groups of people apart.
Methodology
The study used modified t-test and F-statistics to rank SNPs and selected subsets based on classification accuracy using a support vector machine.
Limitations
The study may not account for all possible SNPs and their interactions, and results may vary with different populations.
Participant Demographics
The study involved 210 unrelated samples from four populations: CEU, YRI, JPT, and CHB.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website