Computational trans-omics approach characterised methylomic and transcriptomic involvements and identified novel therapeutic targets for chemoresistance in gastrointestinal cancer stem cells.

We investigated the relationship between methylomic [5-methylation on deoxycytosine to form 5-methylcytosine (5mC)] and transcriptomic information in response to chemotherapeutic 5-fluorouracil (5-FU) exposure and cisplatin (CDDP) administration using the ornithine decarboxylase (ODC) degron-positive cancer stem cell model of gastrointestinal tumour. The quantification of 5mC methylation revealed various alterations in the size distribution and intensity of genomic loci for each patient. To summarise these alterations, we transformed all large volume data into a smooth function and treated the area as a representative value of 5mC methylation. The present computational approach made the methylomic data more accessible to each transcriptional unit and allowed to identify candidate genes, including the tumour necrosis factor receptor-associated factor 4 (TRAF4), as novel therapeutic targets with a strong response to anti-tumour agents, such as 5-FU and CDDP, and whose significance has been confirmed in a mouse model in vivo. The present study showed that 5mC methylation levels are inversely correlated with gene expression in a chemotherapy-resistant stem cell model of gastrointestinal cancer. This mathematical method can be used to simultaneously quantify and identify chemoresistant potential targets in gastrointestinal cancer stem cells.