Copyright©2007 Antony Williams
Following on from my posting last night about Diazonamide A I note that PMR has blogged about this. Peter’s conclusion was that “the blogosphere is starting to emerge as a serious place to look for chemistry.” I have to agree! Just Paul’s one posting on Diazonamide A has resulted in a lot of additional trackbacks!
Â I want to comment on a few of Peter’s comments and offer additional commentary. Before proceeding browse his postingÂ then return back here.
1) regarding the structure on PubChem Peter commented “Lovely. I think itâ€™s correct, but itâ€™s not exactly beautiful.” As I commented inÂ my previous posting…there are TWO structures on PubChem. They are DIFFERENT..check the InChI. It is why they are not deduplicated. And those structures ARE very ugly so there is no way to consider the stereochemistry other than downloading the molfile and cleaning it locally hoping there isn’t a problem in the process!
2) Peter comments “So, off to chemspider which is free….<> There are two more structures but both areÂ equally unreadable”. yes absolutely.Â There has been a lot of commentaries about PubChem being the standardÂ on many blogs so we did NOTÂ clean the 2D structures when we loaded the PubChem dataset. TheseÂ coordinates are exactly as downloaded from PubChem. We are working on improving these displaysÂ by working with a collaboratorÂ to source a new CLEANing algorithm.
3) Peter comments “I try to zoom the formula and get a featureless gray square on both IE and Firefox.” There are six structuresÂ in the search. 350562, 4591072, 10472888, 10478902, 17209260, 17212293. The first four show nice ZOOMed structure images. The last two don’t. Why? The database has just been updated to about 18.5 Million structures, an additional 2 million structures in the past couple of weeks. We are presently deduplicating an additional few million structures and will likely cross 20 million structures by the end of the week. The structure IMAGES have not yet been added for the latest additions. They will appear in the near future following the next deduplication process.
4) Peter comments “So I try Jmol (shown right). Now the molecules are three-dimensional but the coordinates in chemspider are those of the 2-D diagram. Personally I regard this as extremely misleading and would NEVER use Jmol for 2D diagrams, but I shanâ€™t pursue this here.”
We’ve commented on this already on this blog. We have been sourcing a 3D optimizer since PubChem does not provide 3D coordinates. We have one now…more comments will come shortly. We probably have over 20 million structures to optimize and, of course, limited resources. So, we’ll get to this when we get to it. We’re leaving Jmol in place until structures are optimized.Â I encourage you all to addÂ yourselves to the Feedburner signin on the blogs to stay informed. We don’t intentionally mislead…
Â 5) Peter comments ”
So I still donâ€™t know what the molecule is. Where else? Perhaps I can use some more abstractsâ€¦And the fourth one on Pubmed hits gold. Itâ€™s from PNAS:” Also:
The comment “I’m sure it’s correct” is interesting. ESPECIALLY since the structure has been revised by Harrans’ group (J. Li, S. Jeong, L. Esser and P. G. Harran, Angew. Chem., Int. Ed.,2001, 40, 4765). Is the paper in PNAS for the original structure or the revised structure? it’s from February of this year so in this case I think Peter is correct. But you get the point …you just have to consider the hexacyclinol debacle to realize that what is printed for a structure is not necessarily correct.
In fact, my own belief is that while the structure is the best representation of the structure lots of errors are made in drawing them – stereocenters are omitted, the structures are drawn in a visually appealing form and sometimes actually misrepresent the structure and cause changes in stereochemistry especially with double bond orientiation. The structure editor used for drawing should be used to generate the details of the stereocenters and they should then be reviewed for accuracy before accepting the structure representation (and generating InChIs etc)
Recently we announced that we had acquired ChemRefer and that we were going to benefit from the closer relationship! However, even now it is possible to use ChemRefer to search articles and so a search for Diazonamide A on ChemRefer found 6 hitsÂ as shown below.
Â Beautiful structures are just a click away. Some are shown below…all supposed to be Diazonamide A.
Anyways…you get the point. PDF files with direct information about Diazonamide A are one click away on ChemSpider through the integrated ChemRefer service. Overall I would say that the structure drawn by Paul Docherty on the TotallySynthetic Blog is CONSISTENT with that given in Simple indole alkaloids and those with a nonrearranged monoterpenoid unit (Nat. Prod. Rep., 2004, 21, 278) – Royal Society of ChemistryÂ so maybe the case is closed. With this information in hand I look forward to a definitive statement from the expert synthetic chemists so that we can make sure the CORRECT structure for Diazonamide A is on ChemSpider.Stumble it!