Recent Advances in Machine Learning Based Prediction of RNA-protein Interactions.

Researchers

Journal

Modalities

Models

Abstract

The interaction between RNA and proteins play critical roles in many biological processes. However, experimentally characterizing the RNA-protein interaction is still challenging, especially in identifying binding sites. Therefore, many efforts have been devoted to develop high quality computational techniques to study the interaction between RNA and protein. In this direction, many important progresses have been made due to the application of novel techniques and strategies. In this review, the recent advances on RNA-protein interaction were summarized in three aspects, including prediction strategies, input features, and datasets. More specifically, the prediction strategies were classified into traditional machine learning, deep learning, and meta-strategy. The input features were organized as sequential, structural, physicochemical, and evolutionary features. Datasets included structure-based, sequence and annotation based, and experiment based databases. At the end of the article, possible future developments were also discussed.
Copyright© Bentham Science Publishers; For any queries, please email at [email protected].