Hey,
Should I generate all tautomers or just use the canonical tautomer when preparing ligands for docking with CDocker algorithm ? Since the canonical tautomer is the one that will most likely be found in the real physical equilibrium, isnt it more intuitive (and less cpu expensive) to discard all but canonical tautomers ?
-Discovery studio 2016 : Cdocker, Ligand prepair.
This is your decision. But, for me, a suitable procedure involves an exhaustive study of this drug-protein interaction, and that means to study each tautomeric form of the drug. If you have any experimental result, you can begin for the most stable tautomer in your experimental condition and evaluate if it could support the experimental data.
You can justify both taking and discarding non-canonical tautomers. Bear in mind, however, that there are many cases where the conformations of ligand interacting with protein/receptors is completely different from its equilibrium structure in solution. See the great study by Perola and Charifson (http://www.ncbi.nlm.nih.gov/pubmed/15115393) who showed that for more than 10% of ligands the bound conformation is more than 9 kcal/mol above the lowest-energy conformation, which is really a lot.
If you want to be safe and can afford it computationally, do also other tautomers.
I have less than 20 ligands so I believe it is computationally feasible to do an exhaustive search through all of them.
Thanks a lot guys
- Badges
- Science topic