Competition Time - Follow Up

VX-509?


So there were two answers posted in comments to the lunchtime competition recently, and a further one sent by email last night. Look at the comments on the post alongside this summary if you are interested.

The answer is that the compound structure (now known by the INN Adelatinib) is a JAK inhibitor and is almost certainly VX-509.

Everyone seemed to follow a similar strategy, and I've paraphrased this process here. This is quite general for the case of a known structure, unknown research code case (the reverse though is a lot more common - known research code, unknown/proprietary structure). For the latter, look at some of the excellent guidance over at our buddy Chris Southan's blog.

  • Convert the image with OSRA - the CH3 was mildly misinterpreted for me, but easy to fix.
  • Search PubChem/ChemSpider with the molecular structure. There is a hit in PubChem (http://pubchem.ncbi.nlm.nih.gov/summary/summary.cgi?cid=59422203), and this is from SureChem.
  • Search patent literature - most people seemed to use SureChem
  • Find a patent with the structure in (the patents are filed by Vertex, and the compounds are claimed as JAK inhibitors).
  • Search the web for "Vertex JAK inhibitor" as a google/bing/whatever search string, and you find that Vertex have one (VX-509) in Phase 2a, which would be a classic time to file for a non-proprietary name.

There was a CAS number in the original document (here). I didn't include this in the image yesterday, since it doesn't find anything on google (at the time the document was published). A simple google for CAS numbers is great for many new compounds, and the hits often lead to Far Eastern compound suppliers, who are clearly pretty good at snaffling up interesting compound structures from the public web.

It would also have been trivial to have turned the connection diagram into an InChI key and done a Google search (it didn't find anything, at least not yesterday).

Note a few things -
  • The string "VX-509" does not appear in the patent, and so it is an inference that the structure is VX-509 - in reality it would be really really surprising if it wasn't. But you never know. 
  • The patent literature is a pretty good source of interesting stuff, but it's difficult to get, and there is probably also some lag between patent structures being accessible in other databases and from SureChem - SureChem recently announced that they were depositing structures into these sources, which is great, but this example exemplifies, that you will probably end up doing the search twice - and maybe it's best to go to SureChem first?
Also, for me it was a great example of crowd-sourced public annotation - this will end up in ChEMBL_16.

jpo and George