CODD-Pred: A Web Server for Efficient Target Identification and Bioactivity Prediction of Small Molecules

Area of Science:

• Computational chemistry
• Drug discovery
• Bioinformatics

Background:

• Target identification and bioactivity prediction are crucial for drug discovery.
• Existing methods require improvement in accuracy and efficiency.

Purpose of the Study:

• Introduce CODD-Pred, a web server for predicting drug targets and small molecule bioactivity.
• Provide a tool for rapid evaluation and optimization of drug candidates.

Main Methods:

• Developed a double molecular graph perception (DMGP) framework for target prediction using 646,498 small molecules and 640 human targets.
• Created a multi-model self-validation activity prediction (MSAP) framework for bioactivity quantification of small molecules against 56 disease-related targets.
• Utilized curated datasets from the GOSTAR database.

Main Results:

• The DMGP framework achieved over 80% top-5 accuracy in target prediction, outperforming existing servers.
• The MSAP framework demonstrated accurate bioactivity quantification for small-molecule ligands.
• CODD-Pred offers a dual approach for comprehensive drug design prediction.

Conclusions:

• CODD-Pred serves as a valuable, freely accessible tool for accelerating drug discovery.
• The developed frameworks enhance the prediction of drug targets and bioactivity.
• This resource aids researchers in the evaluation and optimization of potential therapeutics.