Developing a novel unsupervised and normalization-free tool for multi-omics integration and biomarker identification
Biomarker identification
Dissecting biological heterogeneity through biomarkers
We combine proprietary and publicly available state-of-the art approaches for the identification of biological signatures allowing the dissection of the biological heterogeneity of a disease, the identification of samples or patients subgroups, as well as the molecular targets underlying the phenotype of interest
What we do
An innovative approach to biomarker signatures
COSBI developed an innovative data-driven approach to the identification and analysis of molecular biomarkers broadly applicable to different data types, including transcriptional, miRNA, and protein profiles, which we complement with available state-of-the art approaches. Our method is based on the extraction of rank-based signatures for each sample/subject, which are used to classify them according to a phenotype of interest (such as control/disease status or response/non-response) or to assign them to novel classes. Starting from a first implementation of this method (SCUDO), we have developed a novel pipeline (MOUSSE) for rank-based multi-omics integration, which outperforms currently available methods.
Our technology
Case Studies
Biomarker identification