[skodapetr/lbvs-environment] Link Down (Issue #1)


The issue is in my lbvs-environment repository. LBVS stands for ligand-based virtual screening. It is a practice in the field of cheminformatics, a domain in which I earned my Ph.D. The idea is to find ligands that bind to a given molecule. What, why … you ask? Well, it can be one of the early steps in drug design, saving resources and potentially improving people’s lives.

But back to the issues, the repository houses code and a link to benchmarking datasets. I would argue that the link and the datasets are the most important part as of now. I planned to add more datasets, yet I never find time to get it done.

This is where you can help. The objective is to update the repository, the code, and the libraries. Next, we can add more datasets to reflect on recent developments in the area. At the end, we should add a few methods and evaluate them using the datasets.

Yet it is not only about getting things done, but also about being FAIR. FAIR is an acronym that stands for Findable, Accessible, Reusable, and Interoperable. The idea is that the process we use to work with the data and the way we publish the data should comply with certain rules.

Do not expect fancy libraries, complex architecture, or high-level software engineering here. On the other hand, if you find the domain of cheminformatics interesting, you like Python, and you are interested in data engineering or machine learning, this may be the right match for you.