DeepMRMP: A new predictor for multiple types of RNA modification sites using deep learning.

RNA modification plays an indispensable role in the regulation of organisms. RNA modification site prediction offers an insight into diverse cellular processing. Regarding different types of RNA modification site prediction, it is difficult to tell the most relevant feature combinations from a variant of RNA properties. Thereby, the performance of traditional machine learning based predictors relied on the skill of feature engineering. As a data-driven approach, deep learning can detect optimal feature patterns to represent input data. In this study, we developed a predictor for multiple types of RNA modifications method called DeepMRMP (Multiple Types RNA Modification Sites Predictor), which is based on the bidirectional Gated Recurrent Unit (BGRU) and transfer learning. DeepMRMP makes full use of multiple RNA site modification data and correlation among them to build predictor for different types of RNA modification sites. Through 10-fold cross-validation of the RNA sequences of H. sapiens, M. musculus and S. cerevisiae, DeepMRMP acted as a reliable computational tool for identifying N1-methyladenosine (m1A), pseudouridine (Ψ), 5-methylcytosine (m5C) modification sites.

Twist Expression in Circulating Hepatocellular Carcinoma Cells Predicts Metastasis and Prognoses.

Hepatocellular carcinoma (HCC) is one of the leading malignancies worldwide. Enumeration of circulating tumor cells (CTCs) has been demonstrated to be a prognostic indicator in HCC. Twist plays a critical role in metastasis and has been proposed as a biomarker for epithelial-mesenchymal transition (EMT). However, links between the expression of Twist in CTCs and HCC clinical parameters are still unclear. This study aims to evaluate the relationship between Twist expression in CTCs and clinicohistopathological risk factors of HCC. Between June 2015 and July 2017, 80 HCC patients and 10 healthy volunteers were enrolled in this study. CTCs were isolated and analyzed by the optimized CanPatrol™ CTC-enrichment technique. Our analysis showed that Twist+ CTCs were detected in 54 of the 80 (67.5%) HCC patients. The positive ratios of Twist+ CTCs correlated with portal vein tumor thrombi, TNM staging, AFP, cirrhosis, tumor number, tumor size, and microvascular invasion. Meanwhile, the follow-up results of the 33 HCC patients who underwent hepatectomy showed that the positive ratios of Twist+ CTCs were closely correlated with the rate of metastasis or recurrence and the mortality rate. The ROC curve analyses suggested that the prognostic evaluation of Twist+ CTCs outperforms CTCs alone. Twist+ CTCs showed higher expression in Glypican-3 protein. In conclusion, Twist expression in CTCs could serve as a biomarker for evaluating HCC metastasis and prognosis.