From: Josh Vermaas (vermaas2_at_illinois.edu)
Date: Mon Oct 06 2014 - 17:24:32 CDT

Hmm... psfgen doesn't complain if you exclude the residue definitions
where the atomnames begin with a number. This is probably a bug, as
there is logic that if the first element of the ATOM token is a digit,
something needs to happen (charmm_parse_topo_defs.c in the psfgen
source). I don't, however, have a clue as to why the logic is the way it
is, or why the force field decided to have atomnames beginning with
digits. If you need one of the residues in question, you have two
choices: use Charmm to make your psf, or be VERY careful and remove the
digits from the front of atomnames, while making sure they are distinct.
Then psfgen should work again. If you don't need one of the residues
that fail, just delete them and be happy.
-Josh
On 10/06/2014 04:51 PM, Chaya Stern wrote:
> Hi Josh,
>
> I downloaded the latest VMD 1.9.2b1 and it's still not working...
>
> On Mon, Oct 6, 2014 at 5:31 PM, Josh Vermaas <vermaas2_at_illinois.edu
> <mailto:vermaas2_at_illinois.edu>> wrote:
>
> Hi Chaya,
> What's you VMD version? Carbohydrate atomtypes are often longer
> than 4 characters, and in VMD 1.9.1, long atomtypes would also
> cause psfgen to choke (it would truncate them down to 4, which
> would cause problems). That has been fixed in 1.9.2b1 to the best
> of my knowledge, so if you aren't using the latest beta, that is
> likely the issue.
> -Josh Vermaas
>
> On 10/06/2014 04:25 PM, Chaya Stern wrote:
>> Hi Josh,
>>
>> I tried doing that with the toppar_all36_lipid_cholesterol.str
>> and toppar_all36_lipid_detergent.str. It worked for the
>> cholesterol but not detergent. After trying each residue
>> separately in the detergent stream file I found that psfgen only
>> runs into this problem when there are both carb and lipid atom
>> types. However, adding top_all36_carp.rtf doesn't solve the
>> problem. So I really don't know what else to try.
>>
>> On Mon, Oct 6, 2014 at 5:09 PM, Josh Vermaas
>> <vermaas2_at_illinois.edu <mailto:vermaas2_at_illinois.edu>> wrote:
>>
>> Hi Chaya,
>>
>> psfgen and stream files never liked one another very much,
>> usually since there are CHARMM commands (set, read, if
>> statements, etc.) in stream files that psfgen doesn't
>> understand (psfgen expects topology files, not mixed ones).
>> What I do is strip those commands out and make a new "pure"
>> topology file that lacks these commands and the parameters.
>> I'm super suspicious as to what selective stream files worked
>> fine, because this is what a session with an unadulterated
>> toppar file looks like using the most recent beta:
>>
>> psfgen) reading topology file toppar_all36_lipid_cholesterol.str
>>
>> psfgen) Toppar stream file for cholesterol. Stream
>> following reading of
>> psfgen) top_all36_lipid.rtf
>> psfgen) par_all36_lipid.rtf
>> psfgen) top_all36_cgenff.rtf
>> psfgen) and par_all36_cgenff.prm
>> psfgen)
>> psfgen) ERROR! FAILED TO RECOGNIZE SET
>> psfgen) ERROR! FAILED TO RECOGNIZE SET
>> psfgen) ERROR! FAILED TO RECOGNIZE IF
>> psfgen) ERROR! FAILED TO RECOGNIZE READ
>> psfgen) Topology for cholesterol
>> psfgen)
>> psfgen) ERROR! FAILED TO RECOGNIZE 31
>> psfgen) skipping statements at end of file due to end or
>> return statement
>>
>> -Josh
>>
>>
>> On 10/06/2014 03:01 PM, Chaya Stern wrote:
>>> Hello,
>>>
>>> 'm having trouble with psfgen parsing the
>>> toppar_all36_lipid_detergent.str file that I downloaded from
>>> MacKerrel's lab page. I've made no changes to the file and
>>> this is what I get when psf tries to parse it
>>>
>>> psfgen) reading topology file ../topology/top_all36_lipid.rtf
>>>
>>> psfgen) \\\\\\\ CHARMM36 All-Hydrogen Lipid Topology File
>>> ///////
>>> psfgen) All comments and questions should be submitted to the
>>> psfgen) parameter forum at the CHARMM website:
>>> www.charmm.org <http://www.charmm.org>
>>> psfgen)
>>> psfgen) Created by CHARMM version 36 1
>>> psfgen) reading topology file
>>> ../topology/toppar_all36_lipid_detergent.str
>>>
>>> psfgen) Toppar stream file for detergents. Stream
>>> following reading of
>>> psfgen) top_all36_lipid.rtf
>>> psfgen) and
>>> psfgen) par_all36_lipid.rtf
>>> psfgen)
>>> psfgen) reading topology from stream file
>>> psfgen) detergent residues
>>> psfgen)
>>> psfgen) Created by CHARMM version 36 1
>>> psfgen) ERROR! Failed to parse atom statement.
>>> psfgen) ERROR! Failed to parse atom statement.
>>> psfgen) ERROR! Failed to parse atom statement.
>>> psfgen) ERROR! Failed to parse atom statement.
>>> psfgen) ERROR! Failed to parse atom statement.
>>> psfgen) ERROR! Failed to parse atom statement.
>>> psfgen) ERROR! Failed to parse atom statement.
>>> psfgen) ERROR! Failed to parse atom statement.
>>> psfgen) ERROR! Failed to parse atom statement.
>>> psfgen) ERROR! Failed to parse atom statement.
>>> psfgen) ERROR! Failed to parse atom statement.
>>> psfgen) ERROR! Failed to parse atom statement.
>>>
>>> So then I tried adding the top_all36_carb and top_all36_prot
>>> to get all the atom types but the problem persisted. I also
>>> tried parsing other stream files in the stream directory and
>>> there was no problem there. I didn't find any weird special
>>> characters either in the file.
>>>
>>> Thanks,
>>> Chaya
>>
>>
>
>