Selective vs. Nonselective
Oct 25th, 2007 by kinasepro
There are a handful of ways to visualize selectivity among kinase inhibitors after you’ve had them Profiled.
The gold standard has become the Ambit panel:
(from the Nature Biotech paper)
A simple chart will get the job done:
(INNO-406 / NS-187, from Blood 05, 3948)
or a heat map like Amphoras’:
If you prefer a reductionist approach you’ve always been able to take it down to a single number with fold shift from its nearest neighbor.
And now it looks like there may be a statistically more meaningful number with the recent J Med Chem on applying the Gini Coefficient to kinase inhibitor selectivty. In case you missed it, check out the spreadsheet in the supporting info.
When you say “The gold standard has become the Ambit panel” are you referring to the way in which they represent selectivity/potency on the “Manning tree”, or the Ambit method itself? I don’t know that I would trust selectivity data obtained against binding to phage-displayed kinase domains, which seems several steps removed from activity assays against (one would hope) more rigorously characterized soluble kinases.
The visualization method is irrelevant imo… heh… But I do think people have become familiar with that image. From my perspective they are the standard because they have the most kinases and its cheaper to run their panel.
I think everybody knows that the Kd’s don’t make any sense, and that Upstate isn’t far behind on the numbers game.
..and what about kinobeads (Nat. Biotech. 2007, 25(9), 1035-1044)?Cellular level..with all the pros and cons.
But hey, aren’t all the successful kinase inhibitors the less selective ones?
Although there are good reasons for targeting several kinases at the same time (at least in Oncology) I’d avoid touching things like PKA, TAK1, LKB-1, DAPK-1 etc.
But for other indications selectivity is much more essential
Look at the problems of p38 inhibitors
And how will GSK-3 inhibitors, for Alzhimer’s, look on such a diagram?