Background Colon cancer is a commonly worldwide cancer with high morbidity and mortality. Long non-coding RNAs (lncRNAs) are involved in many biological processes and are closely related to the occurrence of colon cancer. Identification of the prognostic signatures of lncRNAs in colon cancer has great significance for its treatment. Methods We first identified the colon cancer-related mRNAs and lncRNAs according to the differential analysis methods using the expression data in TCGA. Then, we performed correlation analysis between the identified mRNAs and lncRNAs by integrating their expression values and secondary structure information to estimate the co-regulatory relationships between the cancer-related mRNAs and lncRNAs. Besides, the competing endogenous RNA regulation network based on co-regulatory relationships was constructed to reveal cancer-related regulatory patterns. Meanwhile, we used traditional regression analysis (univariate Cox analysis, random survival forest analysis, and lasso regression analysis) to screen the cancer-related lncRNAs. Finally, by combining the identified colon cancer-related lncRNAs according to the above analyses, we constructed a risk prognosis model for colon cancer through multivariate Cox analysis and also validated the model in the colon cancer dataset in TCGA cohorts. Results Six lncRNAs were found highly correlated with the overall survival of colon cancer patients, and a risk prognosis model based on them was constructed to predict the overall survival of colon cancer patients. In particular, EVX1-AS, ZNF667-AS1, CTC-428G20.6, and CTC-297N7.9 were first reported to be related to colon cancer by using our model, among which EVX1-AS and ZNF667-AS1 have been predicted to be related to colon cancer in LncRNADisease database. Conclusions This study identified the potential regulatory relationships between lncRNAs and mRNAs by integrating their expression values and secondary structure information and presented a significant 6-lncRNA risk prognosis model to predict the overall survival of colon cancer patients.

No abstract

Motivation Drug-food interactions (DFIs) occur when some constituents of food affect the bioaccessibility or efficacy of the drug by involving in drug pharmacodynamic (PD) and/or pharmacokinetic (PK) processes. Many computational methods have achieved remarkable results in link prediction tasks between biological entities which show the potential of computational methods in discovering novel DFIs. However, there are few computational approaches that pay attention to DFI identification. This is mainly due to the lack of drug-food interaction data. In addition, food is generally made up of a variety of chemical substances. The complexity of food makes it difficult to generate accurate feature representations for food. Therefore, it is urgent to develop effective computational approaches for learning the food feature representation and predicting drug-food interactions. Results In this paper, we first collect drug-food interaction data from DrugBank and PubMed respectively to construct two datasets, named DrugBank-DFI and PubMed-DFI. Based on these two datasets, two DFI networks are constructed. Then, we propose a novel end-to-end graph embedding-based method named DFinder to identify DFIs. DFinder combines node attribute features and topological structure features to learn the representations of drugs and food constituents. In topology space, we adopt a simplified graph convolution network-based method to learn the topological structure features. In feature space, we use a deep neural network to extract attribute features from the original node attributes. The evaluation results indicate that DFinder performs better than other baseline methods. The source code is available at https://github.com/23AIBox/23AIBox-DFinder Supplementary information Supplementary data are available at Bioinformatics online.

No abstract