> I've had a chat with a few people RE: this topic and the > consensus is that to *reliably* generate ribbons, you need > at least 1 atom in the amide plane (carbonyl oxygen/nitrogen > or nitrogen's hydrogen) and the alpha carbon.
Yes. Jmol uses the carbonyl oxygen since hydrogen atoms are not generally found in .pdb files. > I investigated whether other applications are able > to apply ribbons to 1alm.pdb and so far non have (ie. > VMD and sybyl on our SGI machine). OK > I also emailed John Stone at uni of Illinois who is the > software engineer for VMD. He indicated the following > > "With some tricky logic, you can use the cross product > of the vectors between 3 CA atoms to find an orthogonal > vector of course, but there are structures where that > doesn't work very well, which is why it gets tricky > pretty rapidly" That is what I was wondering about. I would certainly work fine in helix and turn regions. > So I think support for those representations that > require generation of a > vector which is orthogonal to the backbone is probably > not possible currently when only provided with CA. Also, > there are not a lot of such incomplete (I think generally > theoretical) structures(impression I got from > everyone I've talked to about this), thus it hasn't been > a priority for implementation in other molecular > visualization packages. > > Hope this helps. I've attached an older PDF on ribbons > algorithms for anyone interested. The URL for the html version is: http://sgce.cbse.uab.edu/carson/papers/ribbons86/ribbons86.html > Does jmol currently use the ribbons 2 algorithm or an > older ribbons algorithm ? I do not know if this paper describes 'ribbons 2' or not. The Jmol algorithm is based upon this same 1986 paper by Carson & Bugg, but is not the same. There is a reference to the aforementioned URL in the Jmol source code ... Strands.java In fact, the only thing that Jmol really uses is the first paragraph in the algorithm description, 'Defining peptide planes'. He describes using the carbonyl oxygen to determine the peptide plane. Differences: Carson/Bugg need to know the protein secondary structure. Within a helix, they translate the guide points away from the helix axis. Jmol does not. Carson/Bugg use a B-spline ... because they had hardware support for it. Jmol uses a hermite curve to smoothly connect the points. Miguel ------------------------------------------------------- This SF.Net email is sponsored by Sleepycat Software Learn developer strategies Cisco, Motorola, Ericsson & Lucent use to deliver higher performing products faster, at low TCO. http://www.sleepycat.com/telcomwpreg.php?From=osdnemail3 _______________________________________________ Jmol-developers mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jmol-developers
