We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Test set bias affects reproducibility of gene signatures.
- Authors
Patil, Prasad; Bachant-Winner, Pierre-Olivier; Haibe-Kains, Benjamin; Leek, Jeffrey T.
- Abstract
Motivation: Prior to applying genomic predictors to clinical samples, the genomic data must be properly normalized to ensure that the test set data are comparable to the data upon which the predictor was trained. The most effective normalization methods depend on data from multiple patients. From a biomedical perspective, this implies that predictions for a single patient may change depending on which other patient samples they are normalized with. This test set bias will occur when any cross-sample normalization is used before clinical prediction. Results: We demonstrate that results from existing gene signatures which rely on normalizing test data may be irreproducible when the patient population changes composition or size using a set of curated, publicly available breast cancer microarray experiments. As an alternative, we examine the use of gene signatures that rely on ranks from the data and show why signatures using rank-based features can avoid test set bias whilemaintaining highly accurate classification, even across platforms.
- Subjects
GENOMICS; STANDARD deviations; DATA transformations (Statistics); GENE expression microarrays; CANCER chemotherapy
- Publication
Bioinformatics, 2015, Vol 31, Issue 14, p2318
- ISSN
1367-4803
- Publication type
Article
- DOI
10.1093/bioinformatics/btv157