Re: solvate replicates water box (2 by 2 by 2) and segments (8X) for no apparent reason

From: Steven Samuel Plotkin (steve_at_physics.ubc.ca)
Date: Mon Mar 03 2008 - 00:26:11 CST

Dear Michel,
Yes I just noticed that and was going to post a correction, but you
beat me to it! :)
Thank you for finding that so fast.

The error arose because I had used the the wrong labels (VMD labels
rather than PDB labels which are off by 1, plus I had a typo for one
residue) for the disulfide patch:
patch DISU PROT:57 PROT:146

patching residues other than a cystine apparently adds a SG atom at
position x,y,z = 0,0,0 associated with that residue, indeed very
strange. Without any error messages I proceeded glibly to my doom...

Thanks again,
-S

L. Michel Espinoza-Fonseca wrote:
> On Sun, Mar 2, 2008 at 10:56 PM, Steven Samuel Plotkin
> <steve_at_physics.ubc.ca> wrote:
>
>> Dear NAMD users,
>>
>
> Dear Steven,
>
>
>> I am having a strange problem that I can't seem to track the source of.
>>
>> If I solvate the files ubq.pdb and ubq.psf ( for files see url links
>> http://www.physics.ubc.ca/~steve/NAMD/ubq.pdb and
>> http://www.physics.ubc.ca/~steve/NAMD/ubq.psf)
>>
>> with the following command in VMD:
>> % solvate ubq.psf ubq.pdb -t 6 -o 5test_wb_ubq
>>
>> I get a nice water box around it, with 6 Angstrom distance from box
>> boundary to protein.
>>
>
> This is okay, as expected.
>
>
>> If I solvate the files 5test_sod.psf and 5test_sod.pdb
>> (for files see url links
>> http://www.physics.ubc.ca/~steve/NAMD/5test_sod.pdb and
>> http://www.physics.ubc.ca/~steve/NAMD/5test_sod.psf)
>>
>> with the command:
>> % solvate 5test_sod.psf 5test_sod.pdb -t 6 -o 5test_wb_sod
>>
>>
>
> The problem is simple. If you take a look at the original structure,
> there is a "floating" atom 128 A away from the protein (it seems that
> is a SG from an Alanine, which is completely wrong!). That's why
> you're getting this huge box of water. I'd suggest you to check the
> structure or use a better template (there are tons of small proteins
> that you can use as a test cases).
>
> Good luck!
> Michel
>
>
>> I get a gigantic water box surrounding the protein. Looking at the output
>> it seems to have replicated the box 2X2X2 for reasons that are unclear to
>> me.
>>
>> The output of the ubq solvation is gives 8053 atoms after solvation. The
>> output of the 5test_sod solvation gives 99265 atoms after solvation- way too
>> much. The distances of the waters are up to 100 Anstroms away.
>>
>> The solvation of 5test_sod does not stop after the first segment like it
>> should. Instead it seems to keep replicating segments 8 times, apparently
>> replicating the unit cell to make a 2x2x2. I don't want it to do this, this
>> takes up enough memory my RAM overflows on my workstation and I get a seg
>> fault. I want a 1x1x1.
>>
>> I can't understand what is different about the psf and pdb files that is
>> making the solvate command do this?
>> I thought it might have been a metal atom inside the protein, but I removed
>> it and still the problem remains.
>>
>> links to the output of my solvate commands are here:
>> http://www.physics.ubc.ca/~steve/NAMD/output_solvate_ubq.txt
>> http://www.physics.ubc.ca/~steve/NAMD/output_solvate_sod.txt
>>
>> links to the pdb and psf files are above in case that helps.
>>
>> Sorry if this is a basic question or if the answer is obvious, any help is
>> greatly appreciated.
>>
>> Steve P.
>>
>>
>>
>>
>
>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:49:18 CST