Ah, the aromatics, of course... I only had time to skim-read it, and I happen to have the table of data here on my computer. Well investigated there Bob, and I'm glad your examples are encouraging. I absolutely understand as regards Jmol and desiring an exact value. I look forward to hearing the command, if you figure it out; I'd like to see how fast it can be calc'd in JavaScript at res. 15 or so as you exemplified.
Dave 2009/9/2 Robert Hanson <[email protected]>: > Sorry, I shouldn't have been so critical. Table 3 has some pretty major > differences for aromatics. In the paper they describe multiple approximation > methods and then point out that different programs that do the actual > molecular volume calculation (as Jmol is doing) produce different results. > They hypothesize that these different results are due to different settings > for VDW radii without really checking it out. When they reparameterize they > get better results, but as far as I can tell it is totally ad hoc. > > I would say that Jmol, when loaded with the same VDW radii using, for > example: > > load xxxxx.mol > {carbon}.vdw = 1.7 > {oxygen}.vdw = 1.52 > > gives the same result as their MacroModel values (Tran). I don't see any > point in comparing values to "TSAR" program without first asking for what > they use for VDW radii -- if any program is giving different values, that's > almost certainly all there is to it. > > I only have a few of these models on hand, but here is what I get: > > > compound Bindi Abraham Jmol (isosurface > volume resolution 20 sasurface 0) > > benzene 80.3 95.9 82.7 > chloromethane 42.0 41.1 42.3 > cyclohexane 102.0 103.8 101.7 > testosterone 292.7 309.4 293.6 (resolution > 15) > > So, except for the aromatics, that's pretty good. I apologize! > > OK, so the benefit of the calculation they do is that you don't need a > structure. Just a molecular formula and some knowledge about the number of > aromatic rings. With Jmol, we have structures. So the first thing I would > say is that there's little point in approximating this way; we can do > better. Obviously it shouldn't take an isosurface calculation to do this. I > think you could get an exact value using a simple Jmol script. It's just a > simple analytical calculation. > > Let me see if I can write that.... > > > Bob > > > > On Wed, Sep 2, 2009 at 2:25 PM, N David Brown <[email protected]> > wrote: >> >> 10-20%? Where's that from, Bob? >> >> Egon, there are two sets of values compared; one is TSAR, a QSAR >> package (I'm just quoting here). The other is values from MacroModel. >> Example values are methanol is approximated to 35 instead of 37, while >> N((CH2)3CH3)4BPh4 is 592 vs 609. That latter value is the max >> difference listed for the MacroModel program. >> >> David >> >> >> >> 2009/9/2 Robert Hanson <[email protected]>: >> > Sorry, it just seems to me that 10-20% error is not very good for such a >> > simple calculation. (Just comparing their two methods.) I think these >> > people back in 2003 just didn't have suitable resources. It's not like >> > it's >> > that difficult a calculation, but it is true that without the right >> > tools, >> > you do have a problem. But I realize now that I didn't match my VDW >> > radii >> > with theirs. (I'll reply to Angel in a minute.) Maybe you can check some >> > of >> > this out and get back to us?. >> > >> > Wed, Sep 2, 2009 at 1:11 PM, N David Brown <[email protected]> >> > wrote: >> >> >> >> Given the 677 compounds tested and the proximity to the TSAR values, I >> >> think that's an odd comment. It's an approximation, true, but for the >> >> drop in computational cost I find the accuracy astounding. >> >> >> >> Thanks for the isosurface command I requested. >> >> >> >> >> >> 2009/9/2 Robert Hanson <[email protected]>: >> >> > So I read that JOC article... I think both their methods are crap. >> >> > Jmol >> >> > should be doing a much better job because it is the ACTUAL area, not >> >> > some >> >> > (gross) approximation. >> >> > >> >> > Bob >> >> > >> >> > >> >> > On Wed, Sep 2, 2009 at 10:58 AM, Robert Hanson <[email protected]> >> >> > wrote: >> >> >> >> >> >> Using Jmol 11.8: >> >> >> >> >> >> isosurface volume sasurface 0 >> >> >> >> >> >> should do it. You might have to increase the resolution for good >> >> >> results: >> >> >> >> >> >> isosurface resolution 10 volume sasurface 0 >> >> >> >> >> >> keep increasing the number until you are satisfied. >> >> >> >> >> >> I'll also check that reference. Maybe we can implement that. >> >> >> >> >> >> Bob >> >> >> >> >> >> >> >> >> On Wed, Sep 2, 2009 at 10:10 AM, N David Brown >> >> >> <[email protected]> >> >> >> wrote: >> >> >>> >> >> >>> Ok, thanks for the info, Angel :o) >> >> >>> >> >> >>> Bob, clarification please? >> >> >>> >> >> >>> I've just found an intriguing paper indeed, "Fast calculation of >> >> >>> van >> >> >>> der Waals volume as a sum of atomic bond contributions and its >> >> >>> application to drug compounds" by Zhao, Abraham and Zissimos (J. >> >> >>> Org. >> >> >>> Chem. 2003, 68, 7368-7373). They describe a method of calculating >> >> >>> the >> >> >>> vdW volume using utterly trivial maths - so simple they actually >> >> >>> test >> >> >>> it in an Excel spreadsheet! Incredibly it seems to work to a high >> >> >>> accuracy; examples they provide show deviation from the MacroModel >> >> >>> calc'd value (the long but accurate calculation method) of at most >> >> >>> ~19 >> >> >>> Angstroms. Very happy they've discovered this! >> >> >>> >> >> >>> David >> >> >>> >> >> >>> >> >> >>> >> >> >>> 2009/9/2 Angel Herráez <[email protected]>: >> >> >>> > David, I remebmer reading of a recent feature in Jmol that offers >> >> >>> > the >> >> >>> > volume inside an isosurface. Maybe that can be extended or >> >> >>> > matched >> >> >>> > to >> >> >>> > a vdW surface (there are many options in the isosurface command) >> >> >>> > >> >> >>> > Surely Bob will tell you, but you can follow that line --and let >> >> >>> > us >> >> >>> > know :-) >> >> >>> > >> >> >>> > >> >> >>> >> >> >>> >> >> >>> >> >> >>> >> >> >>> ------------------------------------------------------------------------------ >> >> >>> Let Crystal Reports handle the reporting - Free Crystal Reports >> >> >>> 2008 >> >> >>> 30-Day >> >> >>> trial. Simplify your report design, integration and deployment - >> >> >>> and >> >> >>> focus on >> >> >>> what you do best, core application coding. Discover what's new with >> >> >>> Crystal Reports now. http://p.sf.net/sfu/bobj-july >> >> >>> _______________________________________________ >> >> >>> Jmol-users mailing list >> >> >>> [email protected] >> >> >>> https://lists.sourceforge.net/lists/listinfo/jmol-users >> >> >> >> >> >> >> >> >> >> >> >> -- >> >> >> Robert M. Hanson >> >> >> Professor of Chemistry >> >> >> St. Olaf College >> >> >> 1520 St. Olaf Ave. >> >> >> Northfield, MN 55057 >> >> >> http://www.stolaf.edu/people/hansonr >> >> >> phone: 507-786-3107 >> >> >> >> >> >> >> >> >> If nature does not answer first what we want, >> >> >> it is better to take what answer we get. >> >> >> >> >> >> -- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900 >> >> > >> >> > >> >> > >> >> > -- >> >> > Robert M. Hanson >> >> > Professor of Chemistry >> >> > St. Olaf College >> >> > 1520 St. Olaf Ave. >> >> > Northfield, MN 55057 >> >> > http://www.stolaf.edu/people/hansonr >> >> > phone: 507-786-3107 >> >> > >> >> > >> >> > If nature does not answer first what we want, >> >> > it is better to take what answer we get. >> >> > >> >> > -- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900 >> >> > >> > >> > >> > >> > -- >> > Robert M. Hanson >> > Professor of Chemistry >> > St. Olaf College >> > 1520 St. Olaf Ave. >> > Northfield, MN 55057 >> > http://www.stolaf.edu/people/hansonr >> > phone: 507-786-3107 >> > >> > >> > If nature does not answer first what we want, >> > it is better to take what answer we get. >> > >> > -- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900 >> > > > > > -- > Robert M. Hanson > Professor of Chemistry > St. Olaf College > 1520 St. Olaf Ave. > Northfield, MN 55057 > http://www.stolaf.edu/people/hansonr > phone: 507-786-3107 > > > If nature does not answer first what we want, > it is better to take what answer we get. > > -- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900 > ------------------------------------------------------------------------------ Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day trial. Simplify your report design, integration and deployment - and focus on what you do best, core application coding. Discover what's new with Crystal Reports now. http://p.sf.net/sfu/bobj-july _______________________________________________ Jmol-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/jmol-users
