[gmx-users] presence of hydrogens

[Jocelyne Vreede]

Dear gromacs users, 

I'm simulating a protein with a covalently bound cofactor, that contains a 
phenyl ring. When I use the united atom approach (as given by the PRODRG 
server) I get results that agree with experiments: the cofactor remains 
buried in the protein, stabilized by a hydrogen bonding network. When the 
hydrogen atoms are present explicitly, the cofactor bends toward the solvent, 
thereby disrupting the internal hydrogen bond network and causing the protein 
to start unfolding. 

I've checked both simulations thoroughly, the only differences are the 
presence of hydrogen atoms on the chromophore. I have changed the carbon 
atoms from united to single carbons (CR61 -> CR6 + HCR etc. ) I have used 
charge distributions that have been published. 

Does anyone have a clue where to look for the cause of this discrepancy?

Thanks in advance, 
Jocelyne Vreede
--
Molecular Microbial Physiology
Unversity of Amsterdam, The Netherlands
vreede at science.uva.nl