A major area of interest in data mining and biomedical research is identification of genes involved in clinical class of interest. A method that has been commonly used is analysis of differentially expressed genes . But these traditional methods fail when applied to genes which have high level of heterogeneity.
To circumvent all these problems Nabavi et al.(2016) have identified a new method which calculates q values for each gene expressed differentially in heterogeneous environment. They have used EMDomics to identify genes involved in causing drug resistance to ovarian cancer. The R package for EMDomics is available at http://www.bioconductor.org/packages/release/bioc/html/EMDomics.html.
Reference:
Nabavi S. et. al. (2016) EMDomics: a robust and powerful method for the identification of gene differentially expressed between heterogenous classes. Bioinformatics 32(4) 533-541