Using PROGRD I got . itp and .gro files for Trilinolein. I want to parametrize the Trilinolein based on berger force field for lipids. I found Gromos 96 53a6 force field extended to include Berger parameters. Can I include my .itp file (from prodrg) to this force fields in a top file? How should I modify the ITP file for this purpose? If it is not possible what is your suggestion?
The .itp file from PRODRG needs to have its contents completely re-done. All atom types and bonded parameters will be wrong. All of the charges are likely to be wrong. You can use it as a scaffold for building a suitable topology, but you will need to spend a considerable amount of time fixing the .itp file. Luckily the molecule is very simple to deal with.
Dear Justin,
Thanks for your answer! Now if I want to use just parameters of gromos53a6, based on one of your papers: warning about charges if we get the itp from prodrg, I still need to fix the charges. Is it still the case for ATP? I have now the itp of triolein from ATP See the molecule attached. if I want to use this file, and keep the rest of parameters, how should I fix charges ? how should I get the correct charges?
Negar
Yes, those charges clearly need correction. The haphazard charges on methylene and methyl groups is the first clue; they should all (by definition in this force field) be zero. If your stated goal is to use Berger parameters for this molecule, then the path forward is clear. Assign the atom types from that force field, zero charges on all methyl and methylene groups, and ester charges are pre-defined. Note that Berger parameters may not be best out there; recent Gromos96 parameter sets from Kukol and others have been shown to be better. But again, the charges are other parameters are prescribed for you. There's no guesswork in that case.
Dear Justin,
I am trying to get correct charges, I guess except the methyl and methylene groups on tails of fatty acid, as you said, the rest are non zero (the part in a circle in attcached file). I looked at the information of rtp file, but when I compare them with some part in itp file of some lipids by Dr. Tieleman, they are not the same!
1. If I want to use gromos 53a6 force field can you please help me with assigning charges? Which ones should be grouped? And what are the assigned charges?
2. In the case that I use BERGER lipid force field, I can use existing parameters for DOPC, but I have to replace the phosphocholine moiety by an oleic chain. I can use charges of some parts but I still need to figure out the charges on CH1-OA-C=O. How should I do that?
3. My final goal is to make interface with water . I want to use the procedure of your tutorial, building biphasic systems. Is it enough to make my itp and include with your extended force field for lipids , the way that you explain in tutorial, and continue like biphasic system tutorial? Am I missing something?
I want to thank you for all the help and very informative tutorials. I appreciate your help.
Negar
1. Each ester group (4 atoms, -CH2-(C=O)-O-) is a charge group. When you refer to the 53A6 parameters, do you mean the original 53A6, or the updated parameters from Kukol? The Berger and Kukol charges here are the same, in any case: 0.5 on CH2, -0.7 on the ester O, 0.8 on the carbonyl C, and -0.6 on the carbonyl O should be adequate.
2. Given the answer to (1), I'd say each of the carbons in the glycerol would be assigned zero charge. There's no real other way to do it if the ester is assigned as in (1).
3. I don't know how this tri-acyl lipid is expected to behave (does it form a bilayer? micelle?), so the biphasic tutorial isn't necessarily the best example to follow, because it assumes the layers are immiscible and regularly structured.
Dear Justin,
Thanks for your helpful answers. I am almost there with the top file based on Berger field !
I want to compare the results of this topology With the one that I got from ATP which is based on gromos 53A6. I need to correct the charges. Can you please help me with the charges that I have to assign to ester groups (for gromos 53a6 force field)?
Regarding your third comment, I am interested in the length scale study which is about three order of magnitude smaller than a droplet. Then I think I can consider the triglycerides as a flat hydrophobic layer in contact with water.
The 53A6 charges are also published; surely you can find them in the literature, but in any case:
0.16 on CH2, -0.36 on ester O, 0.58 on carbonyl C, -0.38 on carbonyl O. Note that 53A6 parameters are not very good for lipids; the Kukol model is significantly better.
Dear Justin,
Building a pop.itp file based on Berger Lipid Parameters I came across your conversation here.
I built the pop.itp file based on the dppc.itp file of your KALP-DPPC tutorial.
My question regards the charges of the carbon in the glycerol and the estergroups as discussed above.
The estergroup does not contain the carbon of the glycerol, but the first of the 16 carbon atoms of the palmitoyl-side-chain, is that correct? Then the glycerol-carbons could be charged zero and the estergroup with the mentioned charges.
Now in the dppc.itp file of the tutorial, the carbon C3 of the glycerol is part of the chargegroup: CH2-O-C=O. And all the carbons of the sidechain are charged with zero.
Is there just one right way to charge the groups or are both ways possible?
I attached my pop.itp file
Thank you in advance!
Anna
The only charged atoms are the four in the ester; your topology looks reasonable enough. Charge groups are approximations and are generally a pretty outdated convention, anyway.
Thanks a lot!
I understand, that the charges are outdated. Nevertheless I'm still curious.
If only the estergroup is charged, it confuses me why in the DPPC file of the tutorial not the estergroup but one carbon of the glycerol builds a chargegroup together with the O doublebond as well as the ester O.
For orientation I am attaching the dppc file.
Here atom 32 LC2, 33 LOS, 34 LC and 35 LO form a chargegroup. Which means the carbon of the glycerol atom 32 LC2, is not charged zero.
The ester contains the atoms 33 LOS, 34 LC, 35 LO and 36 LP2, as I understand it, and does not build a chargegroup in the dppc.itp file.
Since I would like to give reason why I charged the groups the way I did, your help would be very much appreciated!
Charge groups are approximations of a small group that are used solely for the purpose of neighbor searching; they do not inherently have any connection with easily definable chemical groups, but often do. Charge groups generally sum to zero but this is not necessary with PME. Whatever is in the DPPC topology is correct; if this does not match what you have done then I apologize for any confusion but I am going based on assumptions since I do not have a coordinate file that matches your molecule so I can't verify what the atoms actually are.