Abstract
Colorectal cancer (CRC) represents one of the most common malignancies with high morbidity worldwide. RNA methylation (m6A) has been considered to tremendously contribute to cancer initiation and progression since its first discovery. In this study, we comprehensively analyzed associations between mRNA expressions of m6A regulators and CRC tumor samples' epidemiologic information from the Cancer Genome Atlas (TCGA). Multivariate Cox proportional hazard model was applied to screening of m6A regulators whose mRNA expressions were significantly associated with CRC tumor samples' overall survival (OS) probability and those significant regulators were used for LASSO regression analysis to construct CRC prognosis prediction signature. As a result, two regulators i.e., YTHDC2 and ALKBH5 were picked out in multivariate analysis. CRC prognosis signature was constructed based on those two regulators through which CRC tumor samples with favorable and inferior prognosis could definitely be distinguished independent of potential confounding factors. This study should be helpful for identifying prognostic different CRC patients and guiding therapeutic method selection.