Francesco Pietra schrieb:
First, I must congratulate with you and your staff for this beautiful work. Then, I would like to see in the future a BAKMDL-type global conformational search in molecular mechanics. The basic step in any computational work with natural products, be that semiempirical or quantum mechanical, is to get a reliable list of low-lying conformers.
As far as I know we dont have yet features compareable to BAKMDL, but since we have a confomer library for the different amino acids, you might write such an algorithm. Current work is underway to add further conformer features, but this will still take some time...
Second, I have not yet examined in full the possibilities of your MD. I have seen on internet that there is a python script for simulated annealing. I'll try that (and I'll try also to learn python, which I also need for nwchem). What I expect is that this simulated annealing procedure offers the possibility to sample conformers. The beautiful graphics of ballview could aid grouping similar conformers together and than subject them to BAKMDL exploration, once this is implemented.
The beautifull thing about BALLView is, that you can make use of the full set of BALL features. Therefore you can easily modfiy the existing annealing script to use different force fields, settings or record a set of conformers. Also have a look at the IDE features, that will ease your Python scripting: http://www.ballview.org/Gallery/Movies The video at the bottom.
I have not yet checked if there is a debian amd64 package. The molecules I am interested in (no polymers, though huge number of degrees of freedom, require the power of a 64bit multiprocessor machine). To this regard, is any plan to parallelize ballview for dual-opteron?
The debian source package available on debian-mentors allows you to build a binary package from the source package, including the changes for 64 bit arch. Unfortunately I cant create a native 64 bit binary package, since I dont have a corresponding machine. If you have future questions, independent from the debian package, feel free to mail me or post on the BALL mailing list: http://www.ballview.org/Support/MailingLists Best regards Andreas Moll -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
