A Novel Framework for Improving the Prediction of Disease-Associated MicroRNAs
| Autor: | Wenhe Zhao, Jiawei Luo, Nguyen Hoang Tu |
|---|---|
| Rok vydání: | 2019 |
| Předmět: | |
| Zdroj: | Intelligent Computing Theories and Application ISBN: 9783030269685 ICIC (2) |
| Popis: | Increasing evidences have shown that human complex diseases associate with plenty of miRNAs. Identifying potential associations between miRNAs and diseases provides great insight into studying the pathogenesis of complex diseases and improving drugs. However, most proposed prediction methods may not consider the existence of some impossible interactions in these unknown interactions which can be regard as negative interactions. In this paper, we proposed a framework to improve the prediction for some existing algorithms. The framework mainly consists of three steps, the first we cluster miRNAs and diseases from the given dataset by using k-medoids in order to find the weakly related interactions from unknown interactions as negative interactions. Secondly, we use existing algorithms to calculate the associated score matrix for miRNAs and diseases based on the given dataset. Finally, we combine the calculated scores with the potential negative interactions to get the final correlation scores. We conduct comprehensive experiments including 5-fold cross validation (5-fold CV) and leave-one-out cross validation (LOOCV) to indicate that our framework has some advantages including improving performance and universal applicability over several of prediction methods. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|