How would you tell if something is best studied?

How do you define «best studied», what threshold(s) do you use to define «this is good enough»? Hunting for multiple X-ray snapshots of one reaction/type of reaction with enzymes reads like establishing a trajectory like Bürgi & Dunitz pioneered (the later recently embarked for his last way). So «Time-resolved structural studies of protein reaction dynamics: a smorgasbord of X-ray approaches» by Westenhoff et al. in Acta Cryst. (2010) A66, 207–219 (doi 10.1107/S0108767309054361, open access) may offer an entry about these bigger molecules.

@Buttonwood thanks for the paper reference, which sadly does not discuss any of the kinds of reactions I am looking for, just generally tells why it might not be easy but that it is the right thing. Not sure about all this negativity and going off on the word "best". But I added that comment to my original question.

Do people get crystal structures of the enzyme bound to its actual substrate? I was under the impression that it's easier to get the structure of a complex with a competitive inhibitor.

@GuntherSchadow - I have added a comment on your other question. The comments here are not intended to be negative, they are intended to get to the heart of the matter within the site's guidelines.

Following up on @orthocresol , it is very difficult to determine structures of wild-type enzymes with substrates except at cryo temperatures because the substrates react to products more quickly than any structural data collection method. More often you will see structures with inhibitors that mimic the proposed transition state or substrates bound to inactive variants of the enzyme. Some people generate computational models of the substrate in the wild-type active site and maybe even simulate the reaction, but those structures/simulations don't go to PDB usually.

@Andrew, I'd love to have even a predicted best guess model, but as you say, they don't usually go into the PDB. But where might they go? In the original lactose synthase paper I started with they have a model, but they didn't submit it to PDB and the authors are unreachable to try to get the model they had 20 years ago.

I am going to pound on this completely baseless closing of this question. This is a mere formality and someone with ill intent is digging their heels in. I want this to be escalated. This question was closed in error, it is NOT "opinion based" and should be re-opened. I will go all the way to the pope of StackExchange if I have to, because I detest this behavior of closing reasonable stuff on formalities, especially if these formalities are completely baseless nit-pickery.

@ToddMinehardt I am hereby pinging you. You are wrong.

@GuntherSchadow - Feel free to escalate to SE mods for resolution.

Unfortunately, the rules here do not lean in the direction of direct justification.

@ToddMinehardt, there is nothing "unfortunate" about any rules, because it is all in your own discretion what you do. And YOU are wrong. Objectively. You can hide behind the rules but you are wrong, there is NOTHING "opinion based" about this question.

Well, I am sorry you feel that way. The community can decide on the vote to reopen.

@GuntherSchadow - I have asked the other mods to address your issue.

In «Protein Databases on the Internet», (Xu, Curr Protoc Mol Biol. 2004 Nov; CHAPTER: Unit–19.4, dx.doi.org/10.1002%2F0471142727.mb1904s68, open access on NIH: ncbi.nlm.nih.gov/pmc/articles/PMC3265122), there is the explicit note that pdb.org does not longer accept model data.

The one about modeled structures, proteinmodelportal.org, closed its doors, too.

But the page refers instad to swissmodel.expasy.org/repository

which advertises itself with «[to be a] collection of annotated 3D protein models generated by automated homology modelling».

and later «Currently the repository contains 2,153,795 models from SWISS-MODEL for UniProtKB targets as well as 172,276 structures from PDB with mapping to UniProtKB.»