ChemSpider Blog

Chemical names generated by chemists can suffer from quality issues. Systematic nomenclature is not easy and software can definitely help…read this paper for example. If systematic names are not correct then either name to structure conversion algorithms or dictionary look-ups may fail. As we work to provide document markup capability and linkage to chemical structures contained within chemistry documents we need to account for these issues with naming.

As an example, see this online publication where even the same name is inconsistent. See below…

AN_interface is now available to edit the name “under the text” and therefore allow it to be converted OR to link it directly to a chemical on ChemSpider. In this way we are not adulterating the original text but now an “incorrect name” or name that is not associated with a structure for display is now available.

The input is easy…click edit in the structure balloon and then either edit the name or link to a CSID. Also, through this dialog you can declare it as a primary compound, secondary or catalyst. We can of course change this/add to this to be reactant, product, solvent etc.

Following the association/correction in this way the CORRECTed name shows up in the Balloon and the resulting structure is shown (and matches what’s in the article). One more step to optimal document markup.

We are adding additional dictionaries to ChemMantis to support linking to external information. Wikipedia is a rich source of information for chemists and we have chosen to connect out to Wikipedia for details about named reactions. Now, when a person marks up a document and highlights a particular named reaction then the link to Wikipedia is used to populate the information balloon on the article. An example is shown on this article on ChemMantis and the balloon is shown below for the Knoevenagel condensation.

As we work on ChemMantis it is clear that we want to expand the integration out to external sources of information as much as possible rather than limit the connectivities to the ChemSpider platform. We have started to build the necessary dictionaries to support bacteria, fungi, viruses etc so it makes sense to connect these up to external resources. As a proof of concept we are using Wikipedia sources to directly feed the “Species Balloons” and have enabled searching of Wikipedia, Google and Entrex directly from the balloon. As an example of the integration we see below the species balloon filled with the leed of the article from Wikipedia for Zymomonas mobilis(click on the thumbnail)

 

From the balloon it is possible to search across Entrez, Google and directly into Wikipedia for more information. For this particular bacterium Entrez gives a list of results as shown below (click on the thumbnail). We are using a similar approach with elements now. Rather than show a “bare element” in a structure balloon (who needs to see Li for Lithium?) we will display the leed text from Wikipedia for that element. The near future will likely see us link to Uniprot and PDB for proteins and out to similar rich sources for other species.