ADL: Ligand too close. AGFR with non-standard cofactor.
matthew.fisher at stcatz.ox.ac.uk
Mon Dec 21 13:51:04 PST 2020
I've got a bit of an unusual question. I'm docking to a Heme protein that's simulated in the Compound I state (i.e. the central iron of the porphyrin is +4 and the axial ligand is a carbonyl oxygen). Unfortunately, with ADFR, we've been recording poses where the substrate carbon is 2.7-2.9 Angstroms away from the ferryl oxygen which, whilst they are recording experimentally observed modes, is within the Van der Waal's radii of the Fe and C (especially when one considers there's a hydrogen bonded to the carbon that would, in a productive mode, be pointing at the oxygen). Is our initial assumption that these are too close to be true valid, or have we catastrophically misunderstood the output? If they are too close, does anyone have any advice on how we can optimise the ADFR/AGFR routine to produce something more realistic?
It's worth noting our ligands are, currently, using merged nonpolar hydrogens. Is this still correct with ADFR or should we be incorporating all hydrogens? Also, do merged hydrogens affect the VdW radius of the carbon atom?
Thanks for reading and for any thoughts on the above issue.
More information about the autodock