verlinde at u.washington.edu
Thu Feb 24 13:21:35 EST 1994
Scientists in the field of protein structure-based inhibitor design
(I am one myself) are convinced that in computro screening of large
structural databases of small molecules can improve dramatically the
chances for discovering inhibitors of enzymes. A good review of the
results obtained with the program DOCK can be found in Kuntz (1992)
Science 257, 1078.
A big problem, however, with in computro screening is that our scoring
functions are pretty poor. As a result many of the 'leads' found by
these methods are inactive, some of the them are milimolar inhibitors,
and still fewer are micromolar inhibitors or better.
My question is now: if one does real screening (not in computro but
in solution!) of 1,000 random small molecules how many mM, how many
micromolar, and how many inactive compounds does one find one average?
In other words, I would like to know by how many orders of magnitude
in computro screening can help the real in solution screening.
Undoubtedly, scientists in pharmaceutical companies must know the answer.
Mind, however, that I am after the success rate for screening against
an enzyme (because that's what we do in the computer) and not screening
of corporate databases against organisms, cell cultures or organs.
I hope there is somebody out there who knows the answer.
Thanks for your cooperation.
verlinde at gouda.bchem.washington.edu
More information about the Xtal-log