Date: Wed, 30 Jul 2003 16:38:52 -0600 (Mountain Daylight Time)
From: "Thomas E. Cheatham, III"
Subject: Re: AMBER: question about ptraj


Yes, you outline two excellent problems with ptraj namely (1) that the
mask parser is funny and very particular with respect to deciphering
mask specifications (i.e. you cannot quote or AND or OR)
and (2) it is often difficult to decipher my cryptic manual writing.

> diffusion :SPM :801-26516 0.2 average DiffSPMKz15
> diffusion :K+ :799 0.2 average DiffKz15

Diffusion only allows specification of a SINGLE mask string. If you
really mean to calculate the diffusion, averaged for all residues named
SPM and residues 801-26516, I think you can specify:

diffusion :SPM,801-26516 0.2 average DiffSPMKz15

A nice way to figure out if a mask string works and is picking what you
think it should be picking, use rdparm and the "checkmask" command

rdparm prmtop

checkmask :SPM,801-26516


> radial RDF_Kz15_SPM 0.05 8.0 :SPM :799
> radial RDF_NVEk_OO 0.05 8.0 :801-26516@O
> radial RDF_MTS_HH 0.05 8.0 :801-26516@H
> radial RDF_MTS_OH 0.05 8.0 :801-26516@* go
>
> I do not understand why my job stops with the message
> diffusion:solvent mask not specified

Now, the above is a BUG. I built the radial function using the diffusion
routine as a prototype; I did not change the error message which SHOULD
read "radial: no solvent mask specified" as your error is occuring not in
the diffusion command but in the radial command...

Beyond this, I cannot see where the error is, nor can I reproduce it.
Which version of ptraj are you using? Looking at the code, the only time you
should get this error is if the mask string is missing (which it appears
not to be) so I am a bit confused... Alternatively, perhaps one of your
atom selections is not selecting any atoms? This would also break it.

An example of a complicated mask strings that works:

:1-100@O,:101-200@H?,:201-299@H1,:A,@CA

key is no spaces, comma's between entries, ...

We will have an updated parser soon that should make life a tad bit
easier (but still support the current syntax).


Thomas E. Cheatham, III (Assistant Professor) College of Pharmacy