New test case for difficult homologs
Hi Massimo,
I’ve been spending some time with Cele Abad-Zapatero, looking at some very challenging data. The first one he sent is probably twinned, in addition to having strong tNCS and almost certainly being merged in the wrong space group! I’m doubtful that we’ll ever be able to solve that one with the existing models, but I could pass that data set on to you at some point. However, the second data set he sent is much better (not twinned, higher resolution, reasonably isotropic), but is still very challenging.
Basically, the protein is a tetramer, and there’s room for 2 tetramers in the a.u. The top 3 hits in HHpred are 41-54% identical, and you can see from these already that there is substantial flexibility in how the monomers are arranged to form the tetramer. It’s possible to solve the structure using these models, but it’s easy to make the wrong choices and spend a lot of time waiting to see what happens! So I think it’s a good test case for you to try at some point. The questions are whether to use a monomer, dimer or full tetramer as a model, and whether or not to focus on the one structure that has 10% higher sequence identity than the others.
The files you need (HHpred before this structure is ultimately deposited, sequence, dataset) are attached!
Randy