On 20 January 2012 09:39, Róbert Kiss <rk...@mcule.com> wrote: > Hi Noel, > > Thanks for the hint! > > So what do you think, what is the more time-consuming part of the > optimization for large molecules with >10 rotatable bonds? Is that the > systematic search with the first 10 bonds or the subsequent greedy search > for the remaining ones? If the latter, it might help to have an option to > set it even more greedy :) > > A few tests suggest that time is typically below 1 sec up to 10 rot.bonds, > but it can go up to minutes for molecules with a few more rot.bonds.
Each step in the systematic search is followed by the greedy search. So for molecules with more than 10 rotatable bonds, I think the overall time is (2^NROT)*(2*[NROT-10]). An alternative layout algorithm would be a genetic algorithm. > Regards, > Robert > > > > >>-----Original Message----- >>From: Noel O'Boyle [mailto:baoille...@gmail.com] >>Sent: Tuesday, January 17, 2012 7:56 PM >>To: Róbert Kiss >>Cc: openbabel-discuss@lists.sourceforge.net >>Subject: Re: [Open Babel] gen2D parameters, documentation >> >>Check out my revision r4534 at >>http://openbabel.svn.sourceforge.net/viewvc/openbabel/openbabel/trunk/src/m >>cdlutil.cpp?view=log >> >>"mcdlutil.cpp: Speed up structure layout in the general case, by >>carrying out systematic search on the inner N=10 rotbonds, and >>greedily searching the remainder. (The original code carried out >>systematic search on the 'first' N=20 rotbonds leaving the others >>unchanged.)" >> >>Naturally, for a very large number of rotbonds, the code will hit a >>wall. A better strategy might be to set a limit on the greedy search >>after an initial full greedy search. >> >>Any suggestions for improvements, or better still patches, would be >>appreciated. >> >>- Noel >> >>On 17 January 2012 17:41, Róbert Kiss <rk...@mcule.com> wrote: >>> Dear Developers, >>> >>> Is there any documentation available on the --gen2D option of babel? >>> It would be great to know more about how it works. We did some testing >>> and it works well in general. But with some rather complicated >>> molecules it really slows down. I was wondering if there are any >>> parameters that can control the time vs. quality of the 2D structures >>> it generates? >>> >>> Thanks! >>> Robert >>> >>> -- >>> >>> Robert Kiss >>> http://mcule.com >>> >>> ------------------------------------------------------------------------- >>----- >>> Keep Your Developer Skills Current with LearnDevNow! >>> The most comprehensive online learning library for Microsoft developers >>> is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3, >>> Metro Style Apps, more. Free future releases when you subscribe now! >>> http://p.sf.net/sfu/learndevnow-d2d >>> _______________________________________________ >>> OpenBabel-discuss mailing list >>> OpenBabel-discuss@lists.sourceforge.net >>> https://lists.sourceforge.net/lists/listinfo/openbabel-discuss > ------------------------------------------------------------------------------ Try before you buy = See our experts in action! The most comprehensive online learning library for Microsoft developers is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3, Metro Style Apps, more. Free future releases when you subscribe now! http://p.sf.net/sfu/learndevnow-dev2 _______________________________________________ OpenBabel-discuss mailing list OpenBabel-discuss@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/openbabel-discuss
