Yes, originally it WAS surfnet. But I took your suggestion about castp & on my linux station it looked just as bad as the surfnet version in chimera. For some reason, I can't get castp to plot w/in a browser any more - it could be my firewall/antivirus software. Adjusting the cutoff in surfnet didn't make any difference until all you got was a few disconnected footballs, which plotted normally. On the other hand, surfnet works normally on my PC; does seem to be a graphics driver issue. Thanks! Irene
Subject: Re: castp visualization From: meng@cgl.ucsf.edu Date: Wed, 9 Mar 2011 09:39:55 -0800 CC: chimera-users@cgl.ucsf.edu To: einew@hotmail.com
On Mar 9, 2011, at 9:06 AM, Irene Newhouse wrote:
Remember a few days ago, I sent you an image from a Chimera castp visualization & it looked really weird, with strips connecting two apparently disconnected regions? Well, I put up PyMol & its castp viewer & it manages to plot the surface understandably - blackish represents the 'outside' of the cavity... I think my protein is nonetheless somewhat pathological, because if I try to render the same surface in mesh, it swells to occupy nearly the whole molecule, which is not right - mesh & solid should enclose the same volume. However, the solid rendering does reproduce where the substrate goes, and how on earth it gets in there is now my problem:-)
I just tried something else -- I have the new Chimera on my linux station, & that's where I was seeing the odd renderings. On my PC, I have the previous version of Chimera & it shows the same surfaces as PyMol! The image chasurfs.png shows the odd effects [from the current release on my linux station], for only one of the dimer units [I thought maybe the dimer was the issue. No.] Possibly the surface rendering of cavities in the current release of Chimera is correct most of the time, but it looks as if it's not for a protein as odd as this one. For debugging purposes, the PDB is 1X09 [I used Chimera to generate the other dimer unit].
Hi Irene, Given your observations, it may be a graphics driver problem on your linux machine. I.e. the salient difference may be the platform, not the Chimera version. However, wasn't what you sent before from Surfnet, not CASTp? In any case, my understanding is that the same Chimera code is used on the different platforms, and I don't think these tools have been changed for months.
I just tried "open castp:1x09" in the 1.5 release and a recent daily build (v1.6 Feb 24) and looked at a few pockets, and the surfaces seemed fine.
Since I deal with the user interface and not programming or hardware issues, I included the list in case others have suggestions or comments. Best, Elaine