ChemSpider Blog

Those of you who have been reading the blog will likely recall my views the Open Notebook NMR project. I have been trying to establish whether or not the time-consuming and computationally intensive GIAO calculations associated with Open Notebook NMR provide better performance than standard out-of-the-box PC-based desktop computer calculations taking just a few seconds. We didn’t get to examine that hypothesis out of that work but I have been working on that study with a couple of my ex-colleagues at ACD/Labs and hope to report on it in the future. A paper was released this week by Rzepa and Braddock in J Nat Prod. regarding the “Structural Reassignment of Obtusallenes V, VI, and VII by GIAO-Based Density Functional Prediction” doi:10.1021/np0705918

I want to comment on the new feature of Associating a Publication with a structure, in this case associating the J Nat Prod paper with this structure. In this case the structure of Obtusallene V is in the ChemSpider database.

I added the appropriate identifiers using the usual process and then added the publication. To do this simply select “Add Publication” and the following screen will be displayed.

About 60 seconds of work and I had added the article title, authors, DOI and a short decsription regarding the paper. When I submitted it the publication went to the curator for approval, in this case me. When approved then the article is associated with, and linked via DOI, to the structure.

As articles are associated in this way the literature, both open and closed, becomes structure and substructure searchable extending the reach of ChemSpider even further.

I then deposited the CORRECT structure of Obtusallene V as identified by Rzepa and Braddock. The structure is here and scrolling down shows you the annotation regarding it being the correct structure. I removed the name Obtusallene V from the old structure so that now the search on that name takes you to the new and correct structure. This URL is that query: http://www.chemspider.com/Search.aspx?q=obtusallene+V 

This blog post is a perfect example of what we have been promising since the beginning: structure deposition, curation and cleaning up the quality of data.

I have been asked a number of times what I would recommend to look at SDF files. This is specifically because it is possible to download  the information associated with a structure into one of these viewers and store/manipulate/print. The constraint with this question about SDF viewers is generally that the utility needs to be FREE. There are a number of good SDF viewing tools out there especially in the commercial world. Ann Richard from EPA, who runs the DSSTox database, has commented on SDF file viewers here.

I use numerous tools for SDF viewing. My primary desktop tool is ACD/ChemFolder and I use it because I know it inside out since I managed the product for a number of years and ACD/Labs have been gracious enough to provide me with a copy to support the work I am doing to curate Wikipedia chemical structures. ChemSpider has also been given access to ChemAxon and OpenEye tools as a result of the philanthropic nature of these two organizations and their support of both academia and “no login” websites (in the case of ChemAxon).  We also use these tools for the manipulation of SDF files and confirming the results across multiple software products when we see issues with SDF file handling. I don’t have access to Symyx/MDL tools but as the format was defined by MDL in the early days that would also be ideal when there are issues with the SDF file reads. Just today I was told by a company that their SDF files must be fine since ISIS reads them but I have checked them in other tools and they are not readable thus making this exchange format interesting at best to manage the complexities of SDF files from different sources but such is the nature of the domain I’m afraid.

For a free tool I am presently using a tool from Hyleos called the ChemFileBrowser. Under the constraint of Free it is a great tool and is in very active development as detailed on their webpage. It’s stable, fast, gives me what I need to see in terms of SDF file browsing and is priced well. It doesn’t offer searching and is not connected to a drawing tool and those capabilities are available from the commercial products. However, for the distribution and viewing of SDF files I think it’s a great little utility. The screenshot below is for the single structure view but there is also a grid view.

This structure is part of  my viewing of the IUCr structure file. We are presently working through the extraction of chemical names from IUCr articles from 1948 onwards. These are being converted using name-to-structure tools available to us and then manually curated for accuracy…a long and tedious process. These structures will then be put onto the ChemSpider database and links will be formed directly to the associated IUCr articles via DOI. This is a fairly simple, but time-consuming process for organics but very difficult for organometallic complexes for two reasons - support of conversion of complex organometallic names and the representation of those structures in SDF format.

I am about to start work with the NISS’ PowerMV viewer also as an alternative viewing tool for SDF files. I hope to report back on its capabilities should time allow but this will not be for awhile since we are preparing for our shift from ChemSpider Beta to coincide with our first year Anniversary in March.