how to select molecules from the optimize workflow

[adrienchaton]

Hello and thanks for sharing your optimisation model.

It runs through but I am wondering how would you expect selection of the generated molecules for use downstream?

At first I had expected the "survived" molecules to be the ones to select. Looking more into the results I see that all "survived" molecules are from the intermediate generations and that all the "purged" ones correspond to the last generation.

Looking at the statistics of the scores, both "survived" and "purged" ones have molecules either scoring lower or higher than the initial molecule... Would you otherwise select all the molecules with improving score? Or only the "survived" ones with improving score?

Thanks for any hints!