Hi Markus,

This error occurs when the two specified surfaces have no points within the specified distance range. This is a bug -- it should just report the contact area as zero in that case. I've fixed that in tonight’s daily build.

http://plato.cgl.ucsf.edu/trac/chimera/ticket/12966

But there is a second problem, since your commands work on Mac with 64-bit Chimera 1.8.1. The trouble is almost surely that the surface calculation for one of the two sets of atoms failed on 64-bit Windows. Unfortunately the 3rd part surface calculation code we use (MSMS) often fails. Some ideas about how to try to get surfaces are here

​ http://www.rbvi.ucsf.edu/chimera/docs/UsersGuide/surfprobs.html

although often nothing works. The surface calculation fails especially often in 64-bit Windows Chimera. You might try 32-bit Windows Chimera. Our next generation Chimera will have surface calculation that always works, but is not expected out for at least 6 months.

Tom


On Mar 28, 2014, at 11:23 AM, Markus Heller <mheller@cdrd.ca> wrote:

Hi Elaine,
 
Thanks for your reply.  I just filed a bug report on this, as per Tom’s request.
 
I should also mention that I did get some result with your version, but not what I expected: essentially a couple of surface blobs, but nothing like your figure.
 
Anyhow, this seems to be a genuine bug, so I’ll just hang tight and wait for a fix.
 
Cheers
Markus
 
From: Elaine Meng [mailto:meng@cgl.ucsf.edu] 
Sent: Thursday, March 27, 2014 2:39 PM
To: Markus Heller
Cc: chimera-users@cgl.ucsf.edu
Subject: Re: [Chimera-users] Binding pocket surface command script not working
 
Hi Markus,
It is a mystery to me… I tried it in both 1.8.1 and the current pre-1.9 daily build and didn't get the error.  I'd say go ahead and submit a report… we can always close it if it's not reproducible.
 
However, the result I got was somewhat ugly because there were several holes in the pocket.  I then tried a larger cutoff, as in the folowing:
 
open 2v5z
delete :.b
split atoms :sag
surface
measure contactArea #0.2 #0.1 4 slab -.1,0 color tan
measure contactArea #0.2 #0.1 3.99 slab -.11,-.01 color cornflowerblue
~surf #0
~disp; ~ribbon; disp :sag
sop hideDust #1,2 size 1 metric sizeRank
 
which I think got rid of as many of the holes as reasonably possible, but there were still a couple of ugly places that might be surface calculation artifacts:  where the blue shows through the tan near the middle of the attached image, and the little funnel on the upper right.  There is a small opening through the funnel, so I'm not sure whether it's an artifact or the actual shape of the pocket. This general process would be expected to make a better-looking figure for pockets with fewer openings.
 
This pocket could alternatively be considered one large pocket containing both the SAG and FAD.  For example, in a separate Chimera session I tried:
 
open castp:2v5z
 
which reports the two largest pockets as the SAG/FAD pockets of the A and B chains, respectively.  If you are using the "measure contactArea" approach to get the appearance of different inside and outside pocket colors as shown in the image gallery figure, it could be done for this collective pocket with "split atoms :sag,fad" (well, assuming you didn't get that mysterious error again!).  Sorry for the weirdness,
Elaine
----------
Elaine C. Meng, Ph.D. 
UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab
Department of Pharmaceutical Chemistry
University of California, San Francisco

<image001.png>


On Mar 27, 2014, at 1:05 PM, Markus Heller <mheller@cdrd.ca> wrote:

Hello,
 
I’m trying to reproduce the Binding Pocket Surface image from the Gallery with a different pdb file.
 
I can successfully reproduce with the pdb used in the example (2zcp), but not with 2v5z.  Running latest Chimera.
 
Steps to reproduce:
 
open 2v5z
delete :.b
split atoms :sag
surface
measure contactArea #0.2 #0.1 3 slab -.1,0 color tan
 
gives an error:
 
error: Array must be 2-dimensional, got 1-dimensional
 
 
  File "C:\Program Files\Chimera 1.8.1\share\MeasureVolume\__init__.py", line 60, in surface_area
    area = surface_area(varray, tarray)
 
See reply log for Python traceback.
 
Why, and would it be helpful to file a bug report?
 
Thanks and Cheers
Markus
 
--
Markus Heller, Ph.D.
NMR Scientist
CDRD - The Centre for Drug Research and Development
2405 Wesbrook Mall, Fourth Floor | Vancouver, BC  V6T 1Z3 | Main: (604) 827-1147
Direct: (604) 827-1122 | F: (604) 827-1299 | E: mheller@cdrd.ca | www.cdrd.ca
<image007.png>
 
 
 
 

Follow us: <image003.jpg> <image004.jpg> <image005.png> <image006.png>
 
This email and any files transmitted with it are confidential and intended solely for the addressee.  If you are not the named addressee you should not disseminate, distribute, copy, or alter this email.
 
 
_______________________________________________
Chimera-users mailing list
Chimera-users@cgl.ucsf.edu
http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users
 
_______________________________________________
Chimera-users mailing list
Chimera-users@cgl.ucsf.edu
http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users