Q1:
I thoroughly enjoyed reading your recent article regarding allosteric hotspot
detection.

I am interested in using this for my proteins and have started submitting my
queries via the provided web server.

It has been running for quite a while, since thursday/friday and I wondered
whether if this was normal or whether something has gone wrong.

In addition to this, as I am wanting to do this for a few more PDBs, is
there an option for a batch input?

A1:
We’re happy to hear that you enjoyed reading about our work, though we apologize about the issues you’re experiencing w/the server. We are investigating this, but it appears as if there is a load issue (all four of our backend CPUs have been running 24/7 since the paper went online). We’ll send you further updates soon.

In the meantime, however, there are two alternative options, and both would also address the need for batch input with multiple structures (the server itself does not provide an option for batch input):

1) We would be more than happy to run as many structures as you like, and we’d start running your structures as soon as you send me the relevant PDBs. WE can run over 1000 structures if necessary (We’d be running them on Yale’s HPC machines, not on Amazon).

2) All of the source code is available through GitHub (github.com/gersteinlab/STRESS). If you already have MMTK installed, then everything should be ready to run on the PDB files with which you’re working.

Q2:
Thank you for your very helpful message. I am emailing off my different account as the other one is having trouble attaching documents.

It would be great if you could help run the batch query as I haven’t yet got MMTK installed on my computer. In total I have 163 structures and would be very grateful if you could run STRESS for them.

Here is the .tsv file attached which contains the PDB_ID and CHAIN_ID for my structures.

Just to note, in the file i sent you, the first PDB id is 1e50, for some reason google has changed that. I hope this is OK!

A2:
Most of your runs are now finished, and you can access them in the link below. There are also a few notes I should mention:

http://homes.gersteinlab.org/people/dc547/.M_Pang/

1) I noticed that some of your structures are NMR structures (as oppose to x-ray crystal structures). I should mention that I have not tested to STRESS framework for such structures, so I can’t say for sure how well it performs. Also, the big issue is that, since NMR structures are generally given as an ensemble, the question is: when running STRESS, which structure should be used? I took the first model in each NMR structure, but this is somewhat arbitrary. Thus, I would interpret the NMR structures with caution. A list of your NMR structures is pasted here:
1f5y.pdb EXPDTA SOLUTION NMR
1gd5.pdb EXPDTA SOLUTION NMR
1k1g.pdb EXPDTA SOLUTION NMR
1o4x.pdb EXPDTA SOLUTION NMR
1rmj.pdb EXPDTA SOLUTION NMR
1urf.pdb EXPDTA SOLUTION NMR
2cr4.pdb EXPDTA SOLUTION NMR
2dn4.pdb EXPDTA SOLUTION NMR
2e7b.pdb EXPDTA SOLUTION NMR
2edk.pdb EXPDTA SOLUTION NMR
2edl.pdb EXPDTA SOLUTION NMR
2js7.pdb1 EXPDTA SOLUTION NMR
2jwa.pdb EXPDTA SOLUTION NMR
2jzx.pdb1 EXPDTA SOLUTION NMR
2kn6.pdb1 EXPDTA SOLUTION NMR
2l4c.pdb1 EXPDTA SOLUTION NMR

2) In the link above, I’ve also provided a gz file of all the original PDB files used.

3) There was an error when running the surface module on 1m5o. I think that this is a result of the fact that this structure is mostly composed of nucleic acid (HETATMs in PDB records, which are removed prior to processing (MMTK fails on HETATMs)

4) The interior-module is not yet complete for all 147 of your structures. The following 4 structures are still running (I can send you the results once they’re complete):
2ozl
2zw3
3ezz
3hn3

That’s it for now. Please take a look at the output whenever you get a free moment, and let me know what you think.