TY - JOUR T1 - A method for the detection of meaningful and reproducible group signatures from gene expression profiles JF - Journal of Bioinformatics and Computational Biology Y1 - 2011 A1 - Licamele,L. A1 - Getoor, Lise AB - Gene expression microarrays are commonly used to detect the biological signature of a disease or to gain a better understanding of the underlying mechanism of how a group of drugs treat a specific disease. The outcome of such experiments, e.g. the signature, is a list of differentially expressed genes. Reproducibility across independent experiments remains a challenge. We are interested in creating a method that can detect the shared signature of a group of expression profiles, e.g. a group of samples from individuals with the same disease or a group of drugs that treat the same therapeutic indication. We have developed a novel Weighted Influence — Rank of Ranks (WIMRR) method, and we demonstrate its ability to produce both meaningful and reproducible group signatures. VL - 9 CP - 3 ER -