IDENTIFICATION OF NON-CODING RNAs WITH CLINICAL IMPLICATIONS IN MOLECULAR SUBTYPES OF COLORECTAL CANCER. AN IN-SILICO APPROACH.

SAULNIER D; ZWENGER A; CROCE MV; ABBA MC; LACUNZA E

Encuentro; REUNION DE SOCIEDADES DE BIOCIENCIA; 2020

Colorectal cancer (CRC) is a frequently lethal disease with hetero-geneous outcomes and drug responses. Multiomics studies have re-vealed the molecular landscape of CRC, enabling the classificationof patients according to 4 consensus molecular subtypes (CMS1-4).CMS1-immune comprises most tumors with MSI. CMS2-canonicaland CMS3-metabolic both show epithelial characteristics. CMS4comprises the more mesenchymal-like cancers, with poor patientprognosis. CMS1 and CMS4 also show poor response to standardtherapies. Non-coding RNAs (ncRNAs) constitute more than 70% ofthe transcriptome. Two main classes have been largely associatedwith cancer; miRNAs and lncRNAs. Since some of these ncRNAscan be detected in human body fluid and have good specificity andaccessibility, they have been suggested to be used as novel poten-tial biomarkers for CRC diagnosis and prognosis as well as in theprediction of the response to therapy. In this study, we performeda classification of 688 CRC tumors obtained from TCGA into the 4CMS employing the CMScaller bioinformatics tool. We character-ized each subtype according to its main features. We then conduct-ed an exhaustive gene expression profile analysis to identify the topupregulated lncRNAs and mature miRNAs in each subtype (CMSk)compared to the rest of the groups (CMSk-1; pvalue