Ask HN: Insights on Computational Tools for Molecular Modeling?

I've recently delved into the complex field of computational biology and chemistry, specifically in the context of molecular modeling to understand potential biological activity of novel compounds. I'm interested in the available tools for tasks such as molecular modeling, ligand docking, virtual screening, and predictive modeling based on molecular geometry. The field seems to be evolving rapidly. What are the most exciting advances or platforms that you've come across recently? In particular, I'm curious about the idea of 'developing and validating' molecules in a way that mirrors the software suite of tests. Are we close to a reality where we could have a comprehensive suite of open-source or commercial computational tools to assist the entire research and development cycle from ideation to production? Furthermore, are there any comprehensive libraries or platforms that serve as a "one-stop-shop" for these computational needs? I'm aware of big players like Nvidia developing ML libraries for this domain. Finally, given the sensitivity of research data and the typical reluctance to let it leave local networks, what does a typical data pipeline look like in this context? Are there established best practices for setting up a data pipeline and facilitating collaboration in a secure and efficient manner? Any insights, advice, or resources you could share would be greatly appreciated. Thanks!