Thoughts on integrating protein mutations into calculation framework and dashboard

[jchodera]

We will want to consider in our object model how we can account for transformations between compounds and transformations between protein variants (i.e. mutations).

We will also want to consider how we might present the results of computed mutations for various compounds on the dashboard. For example, @dotsdl suggests the Compounds page could show a grid of computed point mutation free energies reminiscent of this plot from https://www.biorxiv.org/content/10.1101/2022.01.26.477860v2, which could show the impact of the mutation on the inhibitor binding affinity (reduces affinity, netural, increases affinity, or not computed). We could also summarize the key mutations and key residues that, when mutated, most strongly ablate affinity.

In most uses, only a subset of these mutations will be computed, and only for a subset of compounds.