Supervised learning-based tagSNP selection for genome-wide disease classifications
2008

Selecting Genetic Markers for Disease Prediction

Sample size: 1224 publication Evidence: moderate

Author Information

Author(s): Liu Qingzhong, Yang Jack, Chen Zhongxue, Yang Mary Qu, Sung Andrew H, Huang Xudong

Primary Institution: New Mexico Institute of Mining and Technology

Hypothesis

Can supervised learning methods improve the selection of SNPs for disease classification?

Conclusion

The proposed methods outperform traditional SNP selection techniques in disease classification.

Supporting Evidence

  • The SRFA method improves classification performance in SNP-disease association studies.
  • Both genetic and environmental variables should be considered in disease predictions.
  • The study evaluated methods against popular techniques like SVMRFE and LOGICFS.

Takeaway

This study shows how to pick the best genetic markers to help predict diseases using smart computer methods.

Methodology

The study developed two feature selection methods, SRFA and SVRFA, and applied them to two complex disease datasets.

Potential Biases

Potential overfitting due to small sample sizes in complex models.

Limitations

The study may be limited by the sample sizes and the complexity of the models used.

Participant Demographics

The study included 614 white male patients and 206 white pre- and post-menopausal females with Myocardial Infarction, matched with controls.

Digital Object Identifier (DOI)

10.1186/1471-2164-9-S1-S6

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication