1. SMILES2DTA: a CNN-based approach for identifying drug candidates and predicting drug-target binding affinity.
- Author
-
Mukit, Hasanul, Hossain, Sayeed, Farabi, Mirza Milan, Chowdhury, Mehrab Zaman, Pritom, Ahmed Iqbal, and Rana, Humayan Kabir
- Subjects
- *
CONVOLUTIONAL neural networks , *DRUG discovery , *TIME complexity , *DRUG repositioning , *AMINO acid sequence - Abstract
Computational approaches can speed up the drug discovery process by predicting drug-target affinity, otherwise it is time-consuming. In this study, we developed a convolutional neural network (CNN)-based model named SMILES2DTA (Simplified Molecular Input Line Entry System to Drug-Target Affinity) for predicting the binding affinity between a drug and a target protein. The model utilizes CNNs to process sequences of both drug SMILES and target proteins. SMILES2DTA generates multiple sequences from a single drug SMILES sequence, validates them based on Lipinski's rule of five, and assesses their binding affinity against a target protein sequence. We evaluated our model using publicly available datasets and compared its performance to state-of-the-art methods. The results showed that SMILES2DTA outperformed the existing methods and demonstrated improved accuracy, mean squared error, and area under the precision-recall curve. SMILES2DTA has the potential to speed up the drug discovery process by reducing the time and cost complexity of identifying effective drugs. [ABSTRACT FROM AUTHOR]
- Published
- 2025
- Full Text
- View/download PDF