ChEMBL

Here's an interesting graph (below) - it is of the gap between being assigned a USAN and being approved (in the US). The data covers the last ten years of US approved drugs, and then, only those for which I had a USAN date available (there are a total of 189 drugs covered in this current analysis). There is a big long tail, so some compounds get assigned their USANs about 30 to 40 years before approval (however, this is a very small fraction). Typically, a launched drug gets assigned its USAN about two to three years before approval. So one inference of this observation is that USANs of today are likely (on average, and if they do indeed get approved) in 2012-2013. Also, for USANs that are 8 years or older, there is little chance (on average) of them still becoming drugs.

It would be interesting to try and convert this into a probability estimate of a USAN becomming a drug, if it hasn't been approved yet. There would also appear to be some interesting patterns in the fraction of approved USANs. If I do any of this today, I'll post the results, otherwise, there is a stack of email for me to reply to.

UPDATE So a little bit of toy maths and modelling would indicate that the probability of a compound with an approved USAN ever becoming approved as a drug is in the range 20-25%, and if a USAN has existed for ten years (or more) and not yet approved as a drug, the future likelihood of approval is ca. 2%. So, if we now take USANs that are ten or more years old and apply these estimates there are about 180-200 new 'old drugs' to be discovered/commercialised. Pretty interesting, eh?

The picture is how I would like to imagine my currently lost suitcase on its return to me. Reality will be somewhat different.