CHARMM-GUI & MarvinSpace

Dear MarvinSpace Team,





I would like to thank you for your helpful support to CHARMM-GUI.


Certainly, there are mutual benefits of improving and adding new features in MarvinSpace. As far as I know, CHARMM-GUI is the only website by which one can calculate electrostatic potential and interactively visualize it on the molecular surface. And, there are many novel features as well.





There are a few features we would like to see in MarvinSpace in near future. And, you probably know all of them.





1. Fix of hydrogen visualization in PDB.





2. A couple of problems in electrostatic potential visualization.


2.1 We could not set the default value freely internally.


2.2 The color is suddenly changed when we try to move the molecule. (This need some explanation)





(1) go to the website http://www.charmm-gui.org,


(2) click “generate input” on the top menu,


(3) click “PBEQ solver” on the left menu,


(4) type “1kdx” in Download PDB file, and click “Next Step”,


(5) continue to click “Next Step” until you see “Solvation Energy:”, and


(6) click “view structure” next to “step2_pbeq.phi80”





You may experience a sudden color change when you rotate the molecule, or when you click the default value and rotate the molecule. As you can also see, the default values shown in the menu bar are not the values that we set internally.





3. Reading a CHARMM trajectory. I know you are considering this and I hope that this can be done. This will provide a powerful tool to CHARMM-GUI to look at so many interesting things on the web.





Since we know that you are so busy, we will be patiently wait for the bug fix or new features.





Thanks you again.


Wonpil

Dear Wonpil,





thank you for the kind words, and I'm sorry for the late answer.





We could reproduce the 2nd problem and it will be corrected by the next patch release (it's about 2 weeks, I will post here then to notify you).


The surface is first colored according to the values you set internally, but the values are not set to the color component on the toolbar.


As soon as you click in it, its value is set on the surface, which causes the sudden color change.





We'll try to get on with the other things as well. Thank you for the patience.





Kind regards,


Judit

Dear Judit,





We are having a problem in visualizing a solvated system. To reproduce the problem, please





(1) go to the website http://www.charmm-gui.org,


(2) click "Input Generator" on the top menu,


(3) click "Solvator" on the left menu,


(4) type "1le0" in Download PDB file, and click "Next Step",


(5) continue to click "Next Step" until you see "System Size", and (it may take a few minutes to complete the solvation)


(6) click "view structure" next to "step2_solvator.pdb" or "step2.1_waterbox.pdb"





Nothing appears and I have to restart my web browser. It is possible that our setting is not optimal for water visualization. It might be better to have a cheap way to see water molecules. You can download the pdb file to test locally.





Thanks,


Wonpil

Dear Wonpil,





we can reproduce the problem and the input files are successfully downloaded to test.





It's Miklos who will help you in this, but I was curious and tried a quick test. When I removed the POT and CLA from the end of the PDB file, it worked well.


I wonder if it's something with the indexing.





We'll get back as soon as we find the solution.





Kind regards,


Judit

Hi Wonpil,





this problem is related to the extended pdb format that you introduced. Formerly your colleague Vidyashankara in his post http://www.chemaxon.com/forum/viewpost8413.html#8413 informed us that the extension column could either be PROA, PROB etc, and could also begin with C in which case it is followed by 3 digits. In this file this precondition does not hold, as they are lines that contain CLA (last lines of the PDB).


What do these lines depict? Perhaps Chlorine ions?





We will fix this problem soon.





Miklos

Miklos,





Then, this might be my fault. I did not double check the communication between Vidya and you. The extended column that CHARMM is using is for segment identification (SEGID) which is similar to CHAIN ID in RCSB PDB. SEGID can be anything within 4 characters.





SEGID CLA is used for Chloride Ions. But, I can also use CION or CL for this. For you information, it is a residue name (RESNAME) that has a certain chemical property. As you can guess now, RESNAME for Chloride Ion is CLA. So, it is also possible that SEGID is the same as RESNAME in CHARMM.





In summary, there should be no such a precondition to have 4 characters for SEGID. Anything within 4 is fine such as CLA, POT, or CAL etc.





Thanks,


Wonpil

Dear Miklos and Judit,





Here, we report a few other problems with small molecules.


(1) hydrogens are still too many (I know you are working on this)


(2) some bonds do not appear in attached bdfd.pdb


(3) visualization failed with attached nadp.pdb


(4) the tail part looks weird in attached dmpc.pdb





Regarding the waterbox, I know that you are working on fixing the ions (or character flexibility in CHARMM SEGID). Even without the ions, we could not get the connectivity in water molecules in the water box (above). Are we doing something wrong with a MSpace setup?





Thanks,


Wonpil

Dear Wonpil,





thanks for the updates. We will investigate all problems you reported and then get back with detailed response. Thank you for your patience.





Regards


Miklos

Dear Wonpil,

Quote:

(2) some bonds do not appear in attached bdfd.pdb

It's fixed now, standard bond lengths for F and P compounds were missing from the internal (unfortunately - at present - wired in) bond length table. (Next release will contain this bug fix.)

Quote:

(3) visualization failed with attached nadp.pdb

That's a bit tricky, atom types in some hetatm records is shifted one character position to the right which is not standard and confuses the pdb record reader. We will fix this problem, should be available in the next release of marvin.

Quote:

(4) the tail part looks weird in attached dmpc.pdb

Atom labels contain 3 digits and thus the atom type is shifted one position left with respect to the PDB standard, and the type of these carbon atoms is not properly recognised, this causes valence errors.


The root of the problem is same as above item (3), so these two will be fixed together.

Quote:

Regarding the waterbox, I know that you are working on fixing the ions (or character flexibility in CHARMM SEGID). Even without the ions, we could not get the connectivity in water molecules in the water box (above). Are we doing something wrong with a MSpace setup?

Yes, is we are aware of various problems with your water molecules, and we are working on these.





Thank you again for the bug reports, we try our bests to resolve these issues.





Regards


Miklos

Dear Miklos,





Thanks for taking care of those problems.


The fixes will greatly help in visualization of small molecules in CHARMM-GUI;


http://www.charmm-gui.org/?doc=archive&lib=csml





When MSpace is able to read a CHARMM trajectory, one can also look at the motions along certain normal modes later.





Fixing of water and ion visualization will also help to look at the solvated system generated by CHARMM-GUI.





Thanks again,


Wonpil

Dear Miklos and Judit,





We are facing a problem in visualizing a big molecule such as PDB ID:2PKR as well as its electrostatic surface. Please use the PDB ID and the PBEQ Solver steps described above to reproduce the problem.





Thanks,


Wonpil

Dear Wonpil,





I tried... but apart from freezing my browser (safari) I did not manage to produce anything visible in MSpace, only the right side selection panel appeared, the draw canvas remained black.


Is this the problem you are referring to? I cannot make it out from your email.





The most simple approach for us would be if you sent us the file(s) that is (are) passed to mspace to be viewed. Can you do that?





Thanks and regards,


Miklos

Dear Miklos,





The freezing is the problem I am facing. (sorry for unclear statement)


You can always download files from CHARMM-GUI.


For example, in this case, you can download "step2_pbeq.pdb" and "step2_pbeq.phi80" for visualization.





I have another question.


Can we view an iso-energetic contour?


Sine people often want to see a surface contour that has (for example) +2 kcal/mol potential energy, it would be great to have such a tool.





Thanks,


Wonpil

Dear Wonpil,





o.k. thanks, it's clear now - first I thought it was only Safari that had difficulties.


I will download all related files and try them locally that helps get better error messages and better insight in mspace.





Thanks, we'll get back soon.





Miklos