normal Parameterizing a molecule containg coplanar rings

  • Sadeghi
  • Sadeghi's Avatar Topic Author
  • Offline
  • Fresh Boarder
More
2 years 8 months ago - 2 years 8 months ago #9050 by Sadeghi
Hi,
I want to coarse grain the molecule (NH2C2H6+)(C5H9N)(C=O)(C6H6)(NHCONH)(C6H6)(C3N3)(C4H8NO)2, where the two benzene rings and the triazinic ring are co-planar and the other pendants are relevantly linear in respect to the co-planar rings.

I mapped my structure to the following CG structure: (SQ1)(3-TC3)(TN5a)(3-TC5)(P4)(3-TC5)(3-TN1a)(2-SN1)2 and put constrains on all the ring bonds. (I kept the carbonyl group in one single bead (TN5a) so the sp2 hybridization would contribute to the co-planarity of the structure).

However after the cg-production run the two benzene rings around the P4 group (urea group) bend a bit and also the (3-TC3) and (TN5a) groups (corresponding to the (C5H9N) ring and the carbonyl group (C=O)) rotate and bend toward the benzene rings, making the structure non-planar. Since the rings are not rigid, I have avoided any virtual sites and dihedrals in my model, but I need to keep the rings co-planar to one another. Any suggestion on how to keep the structure of this flexible molecule co-planar?
Thank you
Last edit: 2 years 8 months ago by Sadeghi.

Please Log in or Create an account to join the conversation.

More
2 years 8 months ago #9051 by riccardo
Hi, can you give the SMILES string or IUPAC name of the molecule? I can't fully understand the molecular structure from the info you gave. Thanks.

Please Log in or Create an account to join the conversation.

  • Sadeghi
  • Sadeghi's Avatar Topic Author
  • Offline
  • Fresh Boarder
More
2 years 8 months ago - 2 years 8 months ago #9052 by Sadeghi
Yes, sure the SMILES string is CN(C)C1CCN(CC1)C(=O)C2=CC=C(C=C2)NC(=O)NC3=CC=C(C=C3)C4=NC(=NC(=N4)N5CCOCC5)N6CCOCC6, however please note that the amine group is ionized, so I've mapped it to a charged bead (SQ1).
And also a minor mistake in my previous message, my final CG structure includes 19 beads as:

(SQ1)((2-TC3)(TN2a))(TN5a)(3-TC5)(P4)(3-TC5)(3-TN1a)(2-SN1)2
Last edit: 2 years 8 months ago by Sadeghi.

Please Log in or Create an account to join the conversation.

More
2 years 8 months ago - 2 years 8 months ago #9054 by riccardo
Thanks. The molecule is rather complex. See here for the bead sizes that I would use: github.com/ricalessandri/Martini3-mappin...main/gedatolisib.pdf . Note that the mapping is by no means unique, so a few things could be done differently. However, it's recommended to model aromatic fragments with T-beads and aliphatic rings with S-beads.

I would also suggest building the CG model in steps. So, for example, starting from the phenyl-urea-phenyl fragment fragment - here some thoughts about how I would do it:
- the phenyl-urea-phenyl fragment: (TC5)3-P4-(TC5)5 seems like a good mapping (I think you mapped it in the same way as I did here: github.com/ricalessandri/Martini3-mappin...main/gedatolisib.pdf ).
- To keep the co-planarity, you need to use improper dihedrals involving phenyl-phenyl, phenyl-urea, and urea-phenyl. Have you tried that?
- Some virtual sites constructed based on the phenyl beads may be also useful to define such dihedrals, but I would try to avoid that first.
Last edit: 2 years 8 months ago by riccardo.

Please Log in or Create an account to join the conversation.

  • Sadeghi
  • Sadeghi's Avatar Topic Author
  • Offline
  • Fresh Boarder
More
2 years 8 months ago - 2 years 8 months ago #9056 by Sadeghi
Thank you very much for you response! My mapping is almost the same as yours, except for the aliphatic ring between the amine group (bead 1) and the carbonyl group (bead 5), which I opted for a 2-1 mapping (beads 2-3-4) to keep the symmetry of a ring for this part of the molecule. But if the 3-1 resolution gives a better rigidity for this part I can change it.

I have copied my CG-model based on the 19 beads representation in the comments of the GitHub page: github.com/ricalessandri/Martini3-mappin...44e0e7985130475c1bcd , I would highly appreciate if you could give some suggestions on the best
improper dihedral selections as non of my selections yielded a co-planar structure.

The angle between the planes of the 2 phenyl rings in the atomistic structure is around 85 deg, while for my CG model I got an angle of 57 deg.

Best
Last edit: 2 years 8 months ago by Sadeghi.

Please Log in or Create an account to join the conversation.

More
2 years 8 months ago #9057 by riccardo
Great, thanks for posting that image. Right, you lose the ring plane with a 3-to-1 mapping, that's true. But here some more background on describing aliphatic rings in Martini 3: cgmartini.nl/index.php/component/kunena/...ohexane-in-martini-3 .

Which improper dihedrals did you try? Using your numbering, my first guess would be 9-11-10-12 (to keep 12 and 9-11-10 on the same plane), 9-6-8-7 (similarly for the other phenyl), and then something like 10-12-8-7 should enable to tune the phenyl-phenyl dihedral.

By the way, maybe it's just an artifact of the snapshot, but why your phenyl rings don't look like having different bond lengths? E.g., distance_11-10 seems << distance_11-12.

Please Log in or Create an account to join the conversation.

  • Sadeghi
  • Sadeghi's Avatar Topic Author
  • Offline
  • Fresh Boarder
More
2 years 8 months ago #9059 by Sadeghi
Thanks Riccardo, the 3-1 mapping for the aliphatic rings totally make sense. Also in my my model, I hadn't considered the 10-12-8-7 improper dihedral. I'll make these changes to my model and will update with the result. Hopefully it will result in the desired co-planar structure.

And yes, as you mentioned its the artifact of the snapshot, otherwise the phenyl rings do all have the same length.

Please Log in or Create an account to join the conversation.

Time to create page: 0.111 seconds