As an add-on to my earlier discussion on the "PharmaBLAST" algorithm, I spoke of modifications pertaining to the PASS algorithm, fill spheres representing C, N, O, etc, and possibly generate structures of different properties from the hydrophobic ones to the hydrophilic ones that can fit into the crevice where the active site is, depending on its properties. The structure that fits into the active site can be extracted and recorded into the database.
Such a resource should have two layers of comparisons. The first layer involves a comparison with available pharmacophores in the database and the second layer involves a comparison with structures fitting into the active site of cytochrome P-450 enzymes. This will allow researchers to know in advance the possible cytochrome p-450 enzymes that their candidate drug or xenobiotic may react with.
Citations
1) Kemp CA, Marechal JD, Sutcliffe MJ. Progress in cytochrome P450 active site modeling. Arch Biochem Biophys. 2005 Jan 15;433(2):361-8.
Subscribe to:
Post Comments (Atom)
2 comments:
Perhaps, but the active sites for each of the PDB structures would have to be determined first. This wouldn't be easy, especially if the structure is novel. Alternatively, we can always assume that every pocket is a potential binding site I guess, but this will probably return lots of meaningless hits.
Dear Anonymous:
Good arguements. I did a paper critique long ago, and drug interaction with a cytochrome P-450 enzyme is an issue. The drug can act as a hapten, which is immunogic and can lead to antibodies being raised against the cytochrome P-450 themselves. As such, I wouldn't call such hits meaningless, but just something to take note of.
Leeder JS, Riley RJ, Cook VA, Spielberg SP. Human anti-cytochrome P450 antibodies in aromatic anticonvulsant-induced hypersensitivity reactions. J Pharmacol Exp Ther. 1992 Oct;263(1):360-7.
Post a Comment