I am working on developing force field for 7-dehydrocholesterol (7DHC). 7DHC differs from cholesterol with an extra double bond in one 6-carbon ring. Thus, 7DHC has C5=C6-C7=C8 instead of cholesterol's C5=C6-C7-C8. Generally parameter set for this conjugate ring structure should be optimized using 1,3-cyclohexadiene, but unfortunately I cannot seem to find 1,3-cyclohexadiene in the most recent charmm parameter set. Can somebody suggest what compound to use to parametrize this conjugate system in 7DHC. Alternatively, I would appreciate if someone can provide complete set of force field parameters for 7DHC, if such exists. That will save lot of time!
I also thought there was a 1,3-cyclohexadiene in CGenFF, but upon closer inspection, there doesn't seem to be any.
Anyhow, as a starting point, you could either use a combination of toppar_all27_lipid_cholesterol.str and toppar_all22_prot_retinol.str (for the conjugated double bonds), or use the CHARMM General Force Field (CGenFF), which I think will be more straightforward. Build 7DHC, being careful to use the right atom types for the conjugated double bonds (CG2DC1 for one double bond and CG2DC2 for the other double bond in the case of CGenFF). Run CHARMM, and see which parameters are missing. Those missing parameters can be filled in by analogy; you'll probably find good analogies in CGenFF. If you're not satisfied with the result, you can start parameterizing model compounds based on which parameters you want to improve.
Warning: while writing this, I discovered some pretty bad atom typing mistakes in the cholesterol model compounds CHL1 and CHNS in CGenFF. Please use the updated CGenFF files attached to this post.
Last edited by Kenno; 07/16/0901:49 AM. Reason: More info.
Thank you for your advice. I will try to put together all the parameters. Meanwhile, I am puzzled that 1,3-cyclohexadiene parameters are not explicitly present in the charmm force field file, since it includes much more elaborate compounds. Do you happen to have it? I have never used Charmm before so it would save me lot of time if I had 1,3-cyclohexadiene parameters to build 7-dehydrocholesterol.
The picture is a bit messed up (you might want to use the [code] [/code] tags in the future), but it look correct. As for the charges, you can just use the "standard" ones: methylene groups are neutral and have +0.09 on the hydrogens; for the neutral double bonded carbons, I believe there exists a similar rule with +.15 on the hydrogens. No improper dihedrals are needed.