Generate new drug candidates from difference scoring index.
De novo, designing a new drug for druggable target was impossible in the past.
Generative Adversarial Network (GAN) is a kind of deep learning algorithm for generating a new object. It uses generator and discriminator to learn features from large datasets and generates a new target from index or conditions.
Atom provides a service that generates the most suitable drug structure from our drug generator. With advanced AI technology, we don’t select the candidates by scoring model by model. We use different models combined to provide objective index such as affinity, ADME, Solubility and more. It generates a new structure from an optimized objective index. This way, we can find the most suitable new drug with specific druggable target.