Re: CCL:starting structure for docking

Dear Rowyna,
 in principle your approach to take the the structure from a complex, minimize it
 and start your docking calculations is justifiable. Of course by using this
 structure and its conformation you introduce a bias, that maybe useful or
 dangerous, depending on the docking algorithm and the protein, which you are
 going to dock into.
 The concerns of your colleague are legitimate as well. Even if the minimization
 may take care of any strange bond lengths, angles and torsions, the ligand
 structures that you extract from protein complexes out of PDB may be too
 distorted for the  modelling program to even extract the correct covalent
 structure in the first place. In that case a pure molecular mechanics
 minimization would not help. So in practice it is often more convenient and
 saver to start with a molecule builder from scratch, and, if you would like to
 introduce the conformational bias that I mentioned above, force it onto the
 conformation of the bound ligand.
 Yours
 Thomas
 >Dear CCLers,
 >
 >I hope you will bear with this question of mine but I wasn't able to agree
 with my colleague about it. I've docking a molecule, citrulline, to a protein
 and I obtained the citrulline molecule from a protein-citrulline complex from
 the PDB database website. I think it's suitable by just extracting the
 citrulline sequence and adding hydrogen to it and minimized the structure later
 on before I start my docking process.
 >My colleague, however, doesn't think it's an appropriate way of getting a
 starting structure for docking and I should use a structure from a molecule
 database (i.e. the cambridge crystallographic database or etc). He says that
 maybe the starting structure I have used may not be correct in terms of bond
 lengths/angle or the torsions. I don't think that is the case, since I minimized
 the structure anyway, so it wouldn't be any major difference if i take it from a
 complex structure or if I build the structure from scratch (like using the
 Biopolymer module from InsightII) or if I obtained it as a single molecule from
 a database.
 >I'm do apologize if this sounds very trivial but I would be grateful for
 some clarification as I know most of you will have experience in such matters.
 >
 >Thank you for your time.
 >
 >Rowyna K.
 >
 --
 ..
 Dr. Thomas Hübner
 T2-Consult,  München
 Phone: +49 (89) 21 58 16 80
 Fax:    +49 (89) 21 58 16 82
 thuebner "-at-" t2-consult.de
 http://www.t2-consult.de