Identifying potential drug-target interactions based on ensemble deep learning

Abstract

IntroductionDrug-target interaction prediction is one important step in drug research and development. Experimental methods are time consuming and laborious.MethodsIn this study, we developed a novel DTI prediction method called EnGDD by combining initial feature acquisition, dimensional reduction, and DTI classification based on Gradient boosting neural network, Deep neural network, and Deep Forest.ResultsEnGDD was compared with seven stat-of-the-art DTI prediction methods (BLM-NII, NRLMF, WNNGIP, NEDTP, DTi2Vec, RoFDT, and MolTrans) on the nuclear receptor, GPCR, ion channel, and enzyme datasets under cross validations on drugs, targets, and drug-target pairs, respectively. EnGDD computed the best recall, accuracy, F1-score, AUC, and AUPR under the majority of conditions, demonstrating its powerful DTI identification performance. EnGDD predicted that D00182 and hsa2099, D07871 and hsa1813, DB00599 and hsa2562, D00002 and hsa10935 have a higher interaction probabilities among unknown drug-target pairs and may be potential DTIs on the four datasets, respectively. In particular, D00002 (Nadide) was identified to interact with hsa10935 (Mitochondrial peroxiredoxin3) whose up-regulation might be used to treat neurodegenerative diseases. Finally, EnGDD was used to find possible drug targets for Parkinson's disease and Alzheimer's disease after confirming its DTI identification performance. The results show that D01277, D04641, and D08969 may be applied to the treatment of Parkinson's disease through targeting hsa1813 (dopamine receptor D2) and D02173, D02558, and D03822 may be the clues of treatment for patients with Alzheimer's disease through targeting hsa5743 (prostaglandinendoperoxide synthase 2). The above prediction results need further biomedical validation.DiscussionWe anticipate that our proposed EnGDD model can help discover potential therapeutic clues for various diseases including neurodegenerative diseases.

CANDI: A Web Server for Predicting Molecular Targets and Pathways of Cannabis-Based Therapeutics

“…Moreover, the recent development of deep learning networks has expanded the scope and improved the predictability of target identi cation from various biological databases that have grown enormously with abundant data on protein-ligand complexes. Deep learning models can effectively analyze large datasets and complex biological networks, making them increasingly valuable in modern drug target identi cation (Askr et al, 2023;Chen et al, 2024;Zeng et al, n.d.;Zhou et al, 2023). DRIFT is one such model that helps map the targets for the compounds using deep learning approaches by integrating neural network architecture to predict the target-compound binding a nity using the Yuel algorithm in the backend (Chirasani et al, 2022;Wang & Dokholyan, 2022).…”

Techniques and Strategies in Drug Design and Discovery

“…The work of Lu Wang et al [19] focused on developing a method relatively similar to PASS in order to identify potential drug-target interactions (DTIs). The use of these computational methods can systematically assess and prioritize the most probable targets for a specific drug [20][21][22][23], facilitating more focused experimental validation and uncovering new therapeutic uses for existing drugs [24,25]. Their study introduced AMMVF-DTI, an innovative end-to-end deep learning model employing an attention mechanism and multiview fusion.…”