Dear Matthew,
> Well, I don't exactly know how small/big a molecule should be to be
> considered medium. To give you an idea the taxol derivative that I am
> working with has 147 atoms without hydrogen atoms added.
I did not follow your questions & just looked at your structure. This
looks like two organic drugs (one is taxol) connected by a linker of
polyethyleneglycol PEG
http://en.wikipedia.org/wiki/Polyethylene_glycol.
If you consider this molecule as a single molecule during charge
derivation the conformation you will get after geometry optimization
will not be representative. We have used PEG linkers of different
lengths and for sure the PEG linker does not stay linear during MD
simulations...
See
http://q4md-forcefieldtools.org/images/cbc.jpg
Instead, I would like to suggest you to use (i) R.E.D.
Server/Ante_R.E.D. 2.0 and (ii) R.E.D. Server/R.E.D. IV and to follow
the building block approach. In this procedure, you split the big
molecule into elementary building blocks (BD). In your case you could
consider at least four BD.
- See
http://q4md-forcefieldtools.org/REDS/
- The tutorial: http://q4md-forcefieldtools.org/Tutorial/Tutorial-3.php
- Example of R.E.DD.B. project with PEG linker of different length;
http://q4md-forcefieldtools.org/REDDB/projects/F-84/
PMID: http://www.ncbi.nlm.nih.gov/pubmed/20549756
The advantage of this BD approach are:
- each BD has a well defined conformation and is not 'randomly' chosen.
- cpu time for geometry optimization is far shorter
- In your case a key point in that you can easily define different
lengths for the linker (polymer of EG)... i.e. different analogs for
your initial/big structure.
The main limitation of this BD approach is that you have to design
this BD and the charge constraints (intra- and inter-molecular charge
constraints in your case) used during the charge fitting step to
generate the _molecular fragments_ used to reconstruct the initial/big
molecule. This is not obvious when one first starts, but we can help
you for that; the approach to follow in your case is quite basic...
See once again
http://q4md-forcefieldtools.org/REDDB/projects/F-84/
you will find all what you need there (obviously your case is not a
glycoconjugate).
You will find other features in R.E.D. Server as well:
- Add hydrogens to your structure
http://q4md-forcefieldtools.org/REDS/faq.php#19
- create a totally new structure
http://q4md-forcefieldtools.org/REDS/faq.php#20
- correct a bad input geometry:
http://q4md-forcefieldtools.org/REDS/faq.php#18
regards, Francois
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Nov 03 2011 - 02:00:03 PDT
