Abstract
RNA-binding proteins (RBPs) have a central role in different biological processes like gene regulation, containing transcription and alternative splicing, and provide necessary useful information for patient care. Therefore, identifying binding sites of the RBPs on RNA is the main research direction to understand the procedure of several biological processes. Through biochemical experiments, the RBPs binding sites identification has produced good results but these biochemical processes and exploratory methods consume time and money. Thus, it is important to propose an efficient computational model for the identification of RBPs binding sites. In this study, we propose an intelligent prediction model for RBPs binding sites using a deep learning approach. The proposed prediction model is called kDeepBind. We first apply the one-hot representation for the input sequence to permit successive convolution layers. To declare the binding hidden information from the recognized sequences, the convolution neural network (CNN) model is applied to automatically learn the abstract features. Then, we apply the k-Gram feature extraction technique and concatenate with CNN features. We evaluate the proposed predictor kDeepBind on verified RBP binding sites that are obtained from CLIP-seq datasets. Evaluation results show that the proposed prediction model obtains better performance in the term of ROC-AUC than comparative methods. It is thus highly expected that kDeepBind model will become a useful tool for academic research on RBPs binding sites prediction of RNA as well as in drug discovery. All training and testing data, learned motifs, and trained models can be downloaded at http://nsclbio.jbnu.ac.kr/tools/kDeepBind/
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.