Add Optimize and Save Partial Charge and Bond Order?

Would it be possible to add functionality to Marvin Sketch/Marvin Space similar to that found in ChemDraw/Chem3D, so that structures could be optimized using, e.g., AM1/COSMO or Mopac2009 and saved in a format that would preserve 3D coordinates, correct bond orders, and partial charges? Thank you.

Hi,

Optimisation of 3D structures can be performed over force fields, no higher level models are supported in the Marvin suite. 

At present the ChemAxon force field and the MMFF94 are available. Would these be useful for you?

 

Regards

Miklos

Dear Miklos,

Thank you for your reply. I have been trying to standardize the creation of ligand files for docking experiments using autodock 4.2.3 or implementations of autodock 4.2.3, for example, using YASARA 10.10.4

I think that autodock 4.2.3 and YASARA's implementation of autodock 4.2.3 use the AMBER03 force field. In addition, YASARA has options available for optimization; I have been using AM1/COSMO, AMBER03, and YASARA's ff called YASARA2. YASARA also has AMBER94; is this the same as MMFF94? What is the ChemAxon force field; what other FF is it most similar to?

When I was using Windows for my molecular modeling, I used ChemDraw/Chem3D to produce optimized structures. However, I found that some modeling software was unstable in Windows and I have switched to Mac OS X for most of my modeling work.

After switching to the Mac OS, I have been creating ligand files by drawing them in Marvin Sketch 5.3.8 and optimizing in YASARA. I save them as mol2 files, which has partial charge information, and I generate pdbqt files using autodock's prepare_ligand4.py script with the -C option, which preserves partial charges found in the mol2 file.

However, when I compare various ways of producing mol2 files and pdbqt files, I find differences in bond orders and partial charges. It would be great if I could find some "standard" files that I could use for "positive controls" so that I could select the method(s) that would yield the "correct" partial charges and bond orders in molecules.

Some of the ligands I am preparing contains nitro groups. These deceptively simple molecules have proved difficult, due in part to the difficulty in rendering the nitro groups correctly or having correct renderings preserved when saving and reopening the file in Marvin Sketch.

Any help you could provide would be greatly appreciated!

Thank you.

Best regards,

Rudy

Hi Rudy,

- MMFF and Amber are different force fields. MMFF is more general and beside generating reasonable geometry it does predict the energy fairly accurately.

- ChemAxon force-field is an enhanced form of the Dreiding force-field which is used in Gaussian to generate initial 3d structures.

- Though our mol2 export and import are not fully Tripos compliant, but from partial charges and bond order point of view there shouldn't be problems. Can you give us some examples (structures) with which you experienced some problems? Also describe the workflow, how did you use ChemAxon tools in the structure preparation procedure, what was the input, what was the actual output and why do you think it is not correct.

- In my opinion partial charges may be correct even if they do not agree in two different files, since the tools that produced those files might use different charge models. However, there shouldn't be significant differences.

We can assist you in using ChemAxon tools, but i I'm afraid we cannot give support for other systems, or file formats that we do not recognise (like pdbqt - however, we may consider to implement this structure file format later).

Regards

Miklos