Comprehensive Survey of Recent Drug Discovery Using Deep Learning.

Area of Science:

• Computational chemistry and cheminformatics
• Artificial intelligence in drug discovery
• Pharmacology and pharmaceutical sciences

Background:

• Traditional drug discovery is time-consuming and expensive.
• Artificial intelligence (AI) and deep learning (DL) offer promising solutions to accelerate the process.
• Key challenges include predicting drug-target interactions (DTI) and designing novel molecular structures.

Purpose of the Study:

• To review recent advancements in deep learning for drug-target interaction (DTI) prediction.
• To summarize deep learning methodologies for de novo drug design.
• To provide a comprehensive overview of drug/protein representations, DL models, and benchmark datasets.

Main Methods:

• Literature review of deep learning applications in DTI prediction and de novo drug design.
• Analysis of various drug and protein representation techniques.
• Survey of commonly used deep learning models, datasets, and tools for training and validation.

Main Results:

• Deep learning models show significant potential in enhancing the accuracy and efficiency of DTI prediction.
• Various DL approaches are effective for generating novel molecular structures with desired properties.
• A comprehensive summary of current DL methodologies, representations, and datasets is presented.

Conclusions:

• Deep learning is revolutionizing drug discovery by addressing critical challenges in DTI prediction and de novo design.
• Further research is needed to overcome existing challenges and fully realize the potential of DL in pharmaceutical development.
• This review provides a valuable resource for researchers in the field of AI-driven drug discovery.