Open Science Strategy for Building Accurate AI Models in Molecular Space

An exhaustive exploration of the entire molecular space is practically impossible, despite the abundance of structural and biophysical data and the tools for generating more data for training AI models. AI algorithms alone are insufficient to build a highly accurate and precise model. Therefore, this article proposes an overall open strategy, which includes open data, open algorithm, open source code, and open AI model. This strategy aims to enhance transparency, reproducibility, and collaboration in the scientific research community. By adopting this open approach, we can easily validate or retrain our models and continuously improve them using both experimental and computational data and tools.