We are most interested in creative fragment merges, but designs will be prioritized by numerous factors including ease of synthesis, and toxicity modeling. Compounds will then be synthesized by Enamine and tested by groups around the world. PostEra will be running machine learning algorithms in the background to triage suggestions and generate synthesis plans to enable a rapid turnaround. You will be informed of the progress of the molecules through the main stages (validation, synthesis and testing).

https://covid.postera.ai/covid

Does the QuChemPedIA work have any significance for this type of thing?
I have no idea of what they are trying to synthesize of course, but maybe you do.

We are currently working on a molecule generator capable of optimizing properties and respecting constraints. One possible application is of course medicine, but the chemists at our university are specialized in materials. In any case, we are not advanced enough for this very specific subject.

Thank you. I was not sure that my question made any sense at all, and I think you answered it properly.
Keep on crunching.

No, your question makes a lot of sense. Our Boinc project doesn't do that kind of calculation, but the generation of molecules is a subject that occupies us a lot for some time now. Some of our competitors are involved in SARS-COV-2 research. Jensen group for instance: https://colab.research.google.com/drive/1dDefhpk2hFEK1N6jFuVltd-f0VkJ7jBZ. The only difference with this group is that we don't have the tools and knowledge to evaluate the generated molecules. With Thomas, we tried to assess the investment needed to participate. It's far too expensive for us.

Little off topic perhaps, but,
Why invest time and energy in CPU bound projects, when GPU projects are heaps faster, and also support (and will continue to increase in supporting) Double Precision?
I'm sure it's interesting to see how CPU projects can be improved upon, but you won't see a 5x improvement, like seen from running a GPU project vs a CPU project; comparing similar priced and similar in power draw CPUs and GPUs

Because all the codes are not efficient on GPU. GPU are not faster CPU, it's not that simple. Our code is belonging to this category.
We use NWChem which is GPU capable, but not for the tasks we want to do... Another software (proprietary) does GPU for our tasks, but the acceleration is really not interesting. Your GPUs would be much more useful on F@H.

Your GPUs would be much more useful on F@H.

Also on GPUGrid (for Nvidia cards). Folding is short at the moment.