Molecular Design for Drug Discovery with the Use of Generative Adversarial Networks
G. Sumathi, P. Sudha Juliet, M. Shafiya Banu, S. Daisylin Anbu Sujitha, S. A. Yuvaraj, S. Naga Nandini Sujatha · 2024
This research study describes how Generative Adversarial Networks (GANs) are used to improve drug discovery molecular design. The goal is to maximize the discovery of potential drug candidates by making use of GANs' ability to produce a wide variety of chemical structures. The goal is to use better molecular design methods to speed up and improve the medication development process. The purpose of using GANs is to make it easier to build new molecular structures with certain features, such more potency and selectivity. This technology has the potential to bypass the drawbacks of conventional drug discovery techniques and speed up the process of exploring chemical space. It demonstrates that GANs have the potential to revolutionize drug development through computational simulations and validation tests. One potential way to find new treatments and solve medical problems that have so far gone unsolved is to include GANs into molecular design processes. Results from Drug Design Data Resource, Kaggle datasets shows that the molecular formula with id Activity-Based Learning (ABL) minimum value starts with 0.055 and maximum of 10, from the sample of 10 compounds the minimum affinity value is 0.075 and maximum of 3.42. In another data surveyed the drug discovery values show the minimum of the molecular weight is 356.3 and maximum is 471.5.