An ab initio lncRNA identification and functional annotation tool based on deep learning

Abstract

Long noncoding RNAs (lncRNAs) play important biological roles and are implicated in human disease. To characterize lncRNAs, both identifying and functionally annotating them are essential to be addressed. Despite the successes either in identifying from transcripts or predicting interacting proteins by a variety of tools, it continues to be interesting to develop novel ab initio methods to accurately identify and annotate lncRNAs. Moreover, a comprehensive construction for lncRNA annotation is desired to facilitate the research in the field. In this dissertation, we propose a novel lncRNA identification and functional annotation tool named LncADeep. For lncRNA identification, LncADeep integrates sequence intrinsic features and homology features into a deep belief network (DBN) of deep learning algorithm and constructs models targeting both full- and partial-length transcripts. For functional annotation, LncADeep predicts a lncRNA’s interacting proteins based on deep neural networks (DNN) of deep learning algorithm, using both sequence and structure information. Furthermore, LncADeep integrates KEGG and Reactome pathway enrichment analysis and functional module detection with the predicted interacting proteins, and provides the enriched pathways and functional modules as functional annotations for lncRNAs. Test results show that LncADeep has outperformed state-of-the-art tools, both for lncRNA identification and for lncRNA-protein interaction prediction, and then presents a functional interpretation. We expect that LncADeep can contribute to identifying and annotating novel lncRNAs, and providing helpful information for biologists.