Scholarlink

<p>Drug discovery relies on the ability to predict drug-target affinity (DTA), which allows for the efficient identification of drug candidates for certain protein targets. Scalability, accuracy, and interpretability are issues that traditional methods must deal with. In order to improve prediction accuracy, this study proposes a sophisticated approach that combines contact map representations with the Triple Pre-Activated Random Residual Planet Convolution Attention Network (Tri-Pre-A2RP-2CAN). The DTA, KIBA, and Davis datasets are the sources of the input data. Preprocessing employs Focal Vision Transformer with a Gabor Filter for feature enhancement. Feature extraction uses a Dual-Aggregation Transformer (DAT) to capture complex molecular and protein patterns. The modeling framework incorporates Tri-Pre-A2RP-2CAN and RCNN, optimized with PACRTAMN architecture and Planet optimization based hyperparameter tuning. This innovative approach achieves 99.9% accuracy, outperforming existing methods in modeling drug-target interactions. It enhances DTA prediction, improves molecular interaction analysis, and optimizes drug discovery processes, offering scalable and interpretable solutions for pharmaceutical advancements.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Predicting drug-target affinity through triple pre-activated random residual planet convolution coupled attention network and contact maps

  • 期刊论文

  • 发表日期：

    2025年10月7日

查看全文
  • M. Sudha,
  • B. Senthilnayaki,
  • K. Padmanaban,
  • L. Guganathan

摘要

Drug discovery relies on the ability to predict drug-target affinity (DTA), which allows for the efficient identification of drug candidates for certain protein targets. Scalability, accuracy, and interpretability are issues that traditional methods must deal with. In order to improve prediction accuracy, this study proposes a sophisticated approach that combines contact map representations with the Triple Pre-Activated Random Residual Planet Convolution Attention Network (Tri-Pre-A2RP-2CAN). The DTA, KIBA, and Davis datasets are the sources of the input data. Preprocessing employs Focal Vision Transformer with a Gabor Filter for feature enhancement. Feature extraction uses a Dual-Aggregation Transformer (DAT) to capture complex molecular and protein patterns. The modeling framework incorporates Tri-Pre-A2RP-2CAN and RCNN, optimized with PACRTAMN architecture and Planet optimization based hyperparameter tuning. This innovative approach achieves 99.9% accuracy, outperforming existing methods in modeling drug-target interactions. It enhances DTA prediction, improves molecular interaction analysis, and optimizes drug discovery processes, offering scalable and interpretable solutions for pharmaceutical advancements.