molconvert mol2 and SYBYL problems

There are a couple things I want to address:





1) molconvert does not handle mol2 files properly:


I'm converting SMILES to mol2. However, for the molconverted @<TRIPOS>ATOM it does not properly generate the atom_type, and does not generate a substructure id, substructure name, or charge; it does not create a @<TRIPOS>SUBSTRUCTURE section at all.





I've found that SYBYL needs all this information to correctly open a mol2 file. If some information is missing it assigns dummy atom values.





In order to fix this, I wrote a short program that would take the mol2 output of molconvert and "fix" it by adding in the required information.





This leads to the second issue:





2) I open my "fixed" mol2 files in SYBYL and have it add hydrogens. Some aromtic carbons get hydrogenated, when they shouldn't be. The mol2 file correctly lists their atom_type as C.ar. On manually changing them to C.2 I get some of the atoms to display the correct number of bonds, but not all.





Does anyone have any experience with something like this? I might be overlooking something really simple.





Attached are the output of molconvert and the "fixed" version of the same file.

We are going to check it ASAP.

I solved the second problem. It had been setting the atom types of the aromatic oxygens as O.ar, which is not a valid sybyl atom type.





That's out of the way, but the first issue still stands with molconvert. My ad hoc solution works for what I've got, if anyone is interested.

Hi,





thanks for telling us these problems related to Mol2 file format support in Marvin. The primary problem is that we have no access to Sybyl, thus the only way to improve the handling of mol2 files is by our users help.








Regarding the specific problems you reported:





1. SUBSTRUCTURE section is generated now, substructure_id=1 and substructure_name="noname", this enhancement will be available in the next release of marvin (4.1.7) which should be available soon.


According to the description of the mol2 format, the atom charge is not a mandatory field. If you experience that it cannot be omitted please tell us and then we will consider a viable solution.





2. We supposed that C.ar simply means an aromatic carbon and the mol2 export has been implemented respectively. However, we are not sure about how Sybyl interprets aromatic systems. In marvin atom #2 has 3 bonds even after the hydrogenation.


It could be very helpful if you sent us some mol2 files that store aromatic rings and ring systems in order to let us understand Sybyl's representation.





We will let you know when 4.1.7 is available for download, hopefully not later than Monday.





In the meantime, it would be beneficial if you could share your ad hoc solution with the community at the designated forum area: http://www.chemaxon.com/forum/forum10.html.





Regards,


Miklos