A simulated SNP experiment indicates a high risk of over-fitting and false positive results when a predictive multiple SNP model is established and tested within the same dataset.

Abstract excerpt

Several relatively small studies have established predictive models for normal tissue radiosensitivity based on multiple SNPs. Even though these models yielded statistically significant results, the models were often inconsistent with each other. This can presumably be attributed to certain metho...

Topics

Share this publication in a Topic to start or enrich a Post.

• Share to Alleles
• Share to Breast Neoplasms
• Share to False Positive Reactions