is a public web tool for automated learning from heterogeneous omics measurements (mRNA, miRNA, methylation) that makes use of prior knowledge (gene interactions, gene sets, miRNA targets). The resulting models and markers match the actual measurements as well as the relationships among biological entities recorded in curated biological databases. The tool provides the principal means for the user-friendly discovery of dedicated models in particular domains. It represents the platform for assembly, development, comparison and eventual dissemination of the methods for joint analysis of omics data.
- Integrated analysis of mRNA, miRNA and epigenetic data.
- Development of predictive models (support vector machines, decision trees, nearest neighbour search) that best distinguish among user-supplied sample classes.
- Detection of markers (biological modules based on prior knowledge such as miRNA regulations, pathways or gene ontology terms).
- Automated utilization of genomic knowledge during model creation and generation of biomarkers.
- Visualization and data understanding in terms of principal component analysis.
- Automated integration of heterogeneous measurement platforms.
- Smooth search and import of expression samples from NCBI GEO as well as the import of user data.
- Free and easy registration and further utilization.
- The results are all yours, the computational burden is all ours.