The objective of this study is to develop an innovative drug discovery analysis platform which utilizes artificial intelligence techniques to address challenges in the field of drug development due to high cost and high failure rate. The proposed platform applies various artificial intelligence techniques throughout the entire process of new drug development to support data-driven decision-making from the selection of initial candidate materials to the prediction of physiological activities. The platform analyzes molecular structures and pharmacological properties using advanced chemical informatics tools such as RDKit, and can quickly and accurately predict potential candidates from large compound libraries using QSAR modeling and virtual screening algorithms based on deep neural networks. Additionally, visualization functions in the platform enable researchers to easily understand complex analysis results by flexibly linking various data sources and computational tools. The platform is implemented in a Google Colab environment, making it easy for researchers to access without the need for additional expensive computing infrastructures, and integration with Python-based core libraries enables large-scale data analytics and efficient model learning. This approach overcomes the limitations of traditional experimental-focused methodologies and resolves the constraints of existing analysis platforms, allowing researchers to utilize more straightforward and accessible solutions.
저자
Sangjin Kim [ Department of Big Data Science, Korea University/Interdisciplinary Program in Biomedical Data Science Convergence, Korea University, Republic of Korea ]
Jai Woo Lee [ Department of Big Data Science, Korea University/Interdisciplinary Program in Biomedical Data Science Convergence, Korea University, Republic of Korea ]
