From: Rich A. <rap...@oc...> - 2005-08-05 02:30:23
|
In reading through the Molfile V2000 spec contained in "CTfile Formats, June 1997), I noticed that bond order = 4 appears to be limited to queries only. Page 2-6 of the spec, under ttt "bond type" states that "Values 4 through 8 are for SSS queries only." Given that third-party software is either constructing or writing a molecule, and not an SSS query, to/from a molfile, is it reasonable to expect to see anything other than "1", "2", or "3" for bond type in the bond block? I ask this question because I've read scattered statements to the contrary: essentially that bond order = 4 in a molfile represents an "aromatic" bond in a concrete molecule such as benzene. Is there consensus on this issue? How do major vendors and Open Source projects interpret this part of the spec? cheers, rich |
From: Geoffrey H. <ge...@ge...> - 2005-08-05 02:35:36
|
On Aug 4, 2005, at 10:30 PM, Rich Apodaca wrote: > I ask this question because I've read scattered statements to the > contrary: > essentially that bond order = 4 in a molfile represents an > "aromatic" bond in > a concrete molecule such as benzene. The current version of the Open Babel code uses "4" for an aromatic bond. Heck, even the old Babel-1.6 does the same (I just checked the code). I know there are folks from MDL on the Open Babel list, along with a few other commercial vendors. Anyone else care to chime in with implementation issues here? Cheers, -Geoff |
From: Peter Murray-R. <pm...@ca...> - 2005-08-05 13:20:17
|
On Thu, 2005-08-04 at 22:35 -0400, Geoffrey Hutchison wrote: > On Aug 4, 2005, at 10:30 PM, Rich Apodaca wrote: > > > I ask this question because I've read scattered statements to the > > contrary: > > essentially that bond order = 4 in a molfile represents an > > "aromatic" bond in > > a concrete molecule such as benzene. > > The current version of the Open Babel code uses "4" for an aromatic > bond. Heck, even the old Babel-1.6 does the same (I just checked the > code). > > I know there are folks from MDL on the Open Babel list, along with a > few other commercial vendors. Anyone else care to chime in with > implementation issues here? We are working with MDL this summer and we'll see if we can get any more definitive answers. I would like to get to a stage where there is an open clear description of vendor semantics. If that doesn't happen, and when there aren't conformance tools, then third parties guess what is meant by a term and implement something which is similar to, but probably differs from the vendor semantics. And, if the venosr's spec say that 4 should only be used for a given purpose then we should strive to re-orient the community to using it only for that purpose. After all OB can generate explicit Kekule bonds to clear the backlog. If chemistry is to progress into the informatics age then we have to be more intolerant of avoidable fuzzy semantics. The molecular in question shows how easily information can be corrupted. Personally I would like to move twoards explicit hydrogens wherever possible as it entriely removes that particular problem. It also makes InChI more tractable. Given that the total explicit storage for the complete annual output of new chemical structures is less than most people have on their personal music or video store it can no longer be claimed that implicit semantics are a virtue. P. > > Cheers, > -Geoff > > > ------------------------------------------------------- > SF.Net email is Sponsored by the Better Software Conference & EXPO > September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices > Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA > Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf > _______________________________________________ > OpenBabel-discuss mailing list > Ope...@li... > https://lists.sourceforge.net/lists/listinfo/openbabel-discuss -- Peter Murray-Rust Unilever Centre for Molecular Informatics, Department of Chemistry, University of Cambridge Cambridge, CB2 1EW, UK Tel: +44-1223-760369 |
From: Miguel <mi...@jm...> - 2005-08-05 02:53:20
|
> In reading through the Molfile V2000 spec contained in =22CTfile Format= s, > June > 1997), I noticed that bond order =3D 4 appears to be limited to queries= > only. There is a newer version of the ctfile.pdf spec dated October 2003 google for 'ctfile.pdf october 2003' > Page 2-6 of the spec, under ttt =22bond type=22 states that =22Values 4= through > 8 > are for SSS queries only.=22 > > Given that third-party software is either constructing or writing a > molecule, > and not an SSS query, to/from a molfile, is it reasonable to expect to = see > anything other than =221=22, =222=22, or =223=22 for bond type in the b= ond block? Yes, other values are possible. > I ask this question because I've read scattered statements to the > contrary: > essentially that bond order =3D 4 in a molfile represents an =22aromati= c=22 bond > in > a concrete molecule such as benzene. > > Is there consensus on this issue? How do major vendors and Open Source > projects interpret this part of the spec? The October 2003 ctfile.pdf specification (page 12) explicitly states tha= t bond order=3D4 represents an aromatic bond. I don't think there is any ambiguity as to what 4 means. The Oct 2003 version also says in the Notes column that values 4-8 are fo= r SSS queries. It is pretty clear to me that this is a typographical error,= and that it *should* say '5-8'. I base this aon the definitions of values= 5 through 8 5 =3D single or double 6 =3D single or aromatic 7 =3D double or aromatic 8 =3D any These make sense for queries ... Jmol supports this and renders such bonds as solid + dotted. Miguel |
From: Rich A. <rap...@oc...> - 2005-08-05 15:17:33
|
On Thu, 4 Aug 2005 22:53:05 -0400 (EDT), Miguel wrote > There is a newer version of the ctfile.pdf spec dated October 2003 > > google for 'ctfile.pdf october 2003' Thanks for the tip, I was able to find it. > The October 2003 ctfile.pdf specification (page 12) explicitly > states that bond order=4 represents an aromatic bond. > > I don't think there is any ambiguity as to what 4 means. > > The Oct 2003 version also says in the Notes column that values 4-8 > are for SSS queries. It is pretty clear to me that this is a > typographical error, and that it *should* say '5-8'. I base this aon > the definitions of values 5 through 8 5 = single or double 6 = > single or aromatic 7 = double or aromatic 8 = any > > These make sense for queries ... Now this is very interesting because the June 1997 version of the spec says exactly the same thing about bond order = 4 as the October 2003 spec, word-for-word. Taken literally, the meaning is that bond order = 4 means aromatic bond, but only in the context of an SSS query. It is in the same category as bond order = 5, 6, 7, and 8. How likely is it that a typo like this would have persisted in a document that has undergone several revisions over the last several years? I suppose it could happen, but it seems unlikely. cheers, rich |
From: Miguel <mi...@jm...> - 2005-08-05 16:54:33
|
>> It is pretty clear to me that this is a >> typographical error, and that it *should* say '5-8'. > > Now this is very interesting because the June 1997 version of the spec > says > exactly the same thing about bond order =3D 4 as the October 2003 spec,= > word-for-word. Taken literally, the meaning is that bond order =3D 4 me= ans > aromatic bond, but only in the context of an SSS query. It is in the sa= me > category as bond order =3D 5, 6, 7, and 8. > > How likely is it that a typo like this would have persisted in a docume= nt > that > has undergone several revisions over the last several years? I suppose = it > could happen, but it seems unlikely. Actually, to me it seems very likely that the same error would live through another revision of the document. Probably no one questioned it ... the Oct 2003 version probably added new things which were reviewed b= y interested parties ... but noone bothered to review the old stuff. I am convinced that: 1. 4 =3D=3D aromatic in the context of a file 2. the reference to '4-8' for SSS queries in the Notes column is a typographical error that should read '5-8' Miguel |
From: Peter Murray-R. <pm...@ca...> - 2005-08-06 17:15:50
|
On Fri, 2005-08-05 at 10:17 -0500, Rich Apodaca wrote: > On Thu, 4 Aug 2005 22:53:05 -0400 (EDT), Miguel wrote > Now this is very interesting because the June 1997 version of the spec says > exactly the same thing about bond order = 4 as the October 2003 spec, > word-for-word. Taken literally, the meaning is that bond order = 4 means > aromatic bond, but only in the context of an SSS query. It is in the same > category as bond order = 5, 6, 7, and 8. > > How likely is it that a typo like this would have persisted in a document that > has undergone several revisions over the last several years? I suppose it > could happen, but it seems unlikely. > I suspect it is quite likely. The semantics are provided by the programs and if the programs do much the same asas before the documentation may well not change. P. -- Peter Murray-Rust Unilever Centre for Molecular Informatics, Department of Chemistry, University of Cambridge Cambridge, CB2 1EW, UK Tel: +44-1223-760369 |