Recently, we proposed a new approach for analyzing omics data [1].Our approach is conceptually distinctive from previous ones.In contrast to previous approaches where molecular data are unified by some mathematical ways to describe outcome data, from "bottom-up" view point, we adopt a "top-down" approach, by focusing on the outcome difference between samples, rather than those between molecular data.The application of our new approach to diabetes progression data revealed a set of molecular functions, all of which were related to the diseases initiated from diabetes [1].Deductively, this finding inspired us to propose a new concept on cell transformation;cells of one disease could transform cells of another disease, named "disease to disease cell transformation" [1].The performance of our approach were examined for protomics data from lung cancer cell lines to find cancer marker [2], and our new concept for cell transformation was embodied for transcriptome data from malignant prostate cancers to find a company drug [3].