[phenixbb] geometry weight and rmsd bonds - angles

[Nigel Moriarty]

Lionel

>>
>> - bond outlier  CAD - OAC   model 1.26 - ideal: ProDRG single 1.36,
>> readyset deloc 1.30,   eLBOW deloc 1.429,   smiles deloc 1.269
>> - bond outlier  CBC - OBD   model 1.30 - ideal: ProDRG deloc  1.23,
>> readyset double 1.23,   eLBOW aromatic 1.768,   smiles double 1.234
>> - angle outlier CAG CAF CAD   model 99 - ideal: ProDRG 111,   readyset
>> 117,   eLBOW 109,   smiles 117
>> - dihedral outlier CBC NBB CAY CAZ   model 2.2 - ideal: ProDRG 180,
>> readyset 131.66,   eLBOW 62.5,   smiles 68
>>
>> It seems to me that chemistry clearly depends on the software used,
>> amazing.
>
> This shouldn't be a huge surprise - the programs use different methods to
> specify (or guess) molecule parameters, some of which are more effective
> than others, and different approximations when optimizing the geometry, most
> of which are optimized for speed rather than theoretical rigor (unless you
> are running quantum chemical calculations like the AM1 optimization in
> eLBOW).  As a general rule, it is very difficult to accurately guess the
> chemistry of a molecule based on a PDB file alone; I suspect that CIFs may
> have similar problems, but Nigel can clarify.

All ligand programs are parameterising the geometry (and chemistry) of
a ligand to use in a refinement.  This procedure relies greatly on the
quality of the input, so you need to give it the best quality input to
get the best parameterisation of the chemistry.  As Nat said, SMILES
strings are a high quality input and you can add value but using AM1
(--opt).  CIF restraints file have all the information for good
parameterisation but can more easily contains errors so are less
reliable. The most unreliable is a PDB input so many programs need to
make assumptions in order to generate restraints.

The differences between eLBOW and ReadySet! are due to different
inputs.  eLBOW is the engine of ligand restraints generation so its
used in ReadySet!, however, different inputs are used because the
model has a set of PDB atoms and a three-letter residue code.  If the
code is in the PDB ligand database, Chemical Components (CC), things
go well because ReadySet! can draw on the information in the CC to
generate the restraints.  If the code in not in the CC or some other
catastrophic mismatch between ligand code and atom names, ReadySet!
uses the only information available, the PDB lines of the ligand.
This is not the best situation so running eLBOW directly is a better
choice.

I'm happy to take a closer look if you send me the files directly.

Nigel

-- 
Nigel W. Moriarty
Building 64R0246B, Physical Biosciences Division
Lawrence Berkeley National Laboratory
Berkeley, CA 94720-8235
Phone : 510-486-5709     Email : NWMoriarty at LBL.gov
Fax   : 510-486-5909       Web  : CCI.LBL.gov