Cheminformatics and data science toolchains
From Gaussian to Gromacs, from RDKit to quantum computing – the history of cheminformatics, which in some cases goes back more than forty years, has produced a wide range of techniques, implementations, and processes. They are still in use today. With upcoming new technologies, we can assemble these varieties more logically.
Taking the perspective of the digital researcher, there is a demand for building pipelines through a wide variety of subject matters like QM, MM, QSAR, and AI (data science).
Another dimension is the classic IT perspective, which addresses processes like the cloud paradigma, versioning, and security, as well as the handling of high-performance computing. So the domain of digital research is a result of merging the process dimensions of chemistry and IT: model building, testing, teleasing and monitoring.
I have compiled an ontology containing hundreds of objects in the intersection of Chemistry and IT. On that, a cluster analysis is performed to find relevant structures. With these structures, we can model processes. Then, for example, different ecosystems can be compared.
Finally, it is fundament in the construction of an artificial intelligence-based stack performing autochemistry.
(Proposal for ACS 2020 Spring / Philadelphia)