Scholarlink

<p>MicroRNAs (miRNAs) play an important role in developing many chronic human diseases and offer a promising target for new treatments. Researchers are exploring how to control miRNAs using small molecules or drugs to manage disease more effectively. However, finding drugs that can specifically target miRNAs is a difficult and complex task. In this study, we present a new method called bBCOGWO to predict drug–miRNA associations. The approach uses a combination of bee colony optimization (BCO) and grey wolf optimization (GWO) to select important features from drug and miRNA data. These selected features are passed through a convolutional neural network to find meaningful patterns, which are then classified using a support vector machine. We tested the performance of our model using 10-fold cross-validation. It achieved a strong average AUC of 0.9924 and an AUPR of 0.9956. These results show that bBCOGWO can be a useful tool for miRNA-based drug discovery and may support the development of more accurate and personalized treatments. This method can also help us better understand how miRNAs work in chronic diseases.</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–miRNA relationships in chronic diseases using meta-heuristic optimization and convolutional neural networks

  • 期刊论文

  • 发表日期：

    2025年6月13日

查看全文
  • Shivani Saxena,
  • Ahsan Z. Rizvi

摘要

MicroRNAs (miRNAs) play an important role in developing many chronic human diseases and offer a promising target for new treatments. Researchers are exploring how to control miRNAs using small molecules or drugs to manage disease more effectively. However, finding drugs that can specifically target miRNAs is a difficult and complex task. In this study, we present a new method called bBCOGWO to predict drug–miRNA associations. The approach uses a combination of bee colony optimization (BCO) and grey wolf optimization (GWO) to select important features from drug and miRNA data. These selected features are passed through a convolutional neural network to find meaningful patterns, which are then classified using a support vector machine. We tested the performance of our model using 10-fold cross-validation. It achieved a strong average AUC of 0.9924 and an AUPR of 0.9956. These results show that bBCOGWO can be a useful tool for miRNA-based drug discovery and may support the development of more accurate and personalized treatments. This method can also help us better understand how miRNAs work in chronic diseases.