Inconsistent "Inconsistent" structures?

I ran the majorms plugin on every MetaCyc mol file, and found there was two structures for which the plugin stated that they were inconsistent, but the msketch program didn't report any problems when I loaded these structures.  I've attached the files.

CPD-12945.mol --> http://metacyc.org/META/NEW-IMAGE?object=CPD-12945

CPD0-1392.mol --> http://metacyc.org/META/NEW-IMAGE?object=CPD0-1392

Hi,

Please let me know which  version of plugin you've  run.

Thanks.

 

Jozsi

I downloaded the latest MarvinBeans version 5.10.2 to perform these calculations, but I'm unsure as to whether the majorms plugin has its own version number, and how to get it.

Hi,

Ok, I just wanted to know the version number of your MarvinBeans.

Thanks.

Jozsi

Hi,

Please let me know  the pH value.

Do you use tautomer option?

Jozsi

This is how I run it:

cxcalc -N hi majorms -H 7 -f mol:-a <mol file>

I found a third mol file for which I have the same problem, the plugin reports an inconsistent structure, and msketch doesn't.  Attached.

Hi,

 

This a damaged input molecule since it has a valence error on the nitrogen atom. I've attached a figure about it.

This is why the "inconsistent molecule" message  generation is justified.

We investigate what may be the problem wtih your msketch application and go back soon.

Jozsi

Actually, I should add that I went to change the charge on that nitrogen to +1, to resolve the inconsistency, but it was already "checked" to be +1.

 

There were a couple of things that was different this time:

 

1) Usually with a valence error, msketch highlights the offending atom with a red circle.  With this Nitrogen, it just underlined it with a red line.

2) When the file is opened, the Nitrogen does not show a "+" next to it, hence the inconsistency, but if you go to change the charge, the pop-up menu already has a check next to +1, but after you look at that, the "+" then appears next to the Nitrogen.

Update, I can confirm that once I've got the Nitrogen to show the "+" sign, and save the file, then there's no longer and inconsistency problem, so this is more of an MSketch problem really.

Hi,

 

Yes, I remember such kind of bug was fixed recently. In the 5.11 version I could not reproduce this bug. 

I assigned your bug report to the responsible developerr team.  Someone  will respond you soon.

Jozsi

Dear Sam,

Here are my comments about your files:

CPD0-1392.mol
Cisplatin is a coordination complex and please be careful with the bonds. In MarvinSketch you can see valence errors on the nitrogen atoms, which is correct, because N has 3 valences.
The 4th line (bond) which connects the nitrogen to Pt is a coordinate bond, and if you correct it in Marvin to coordinate bonds, you will not get valence errors.
Please find this file attached. (It is exported from MarvinSketch in mol format.)

CPD-7846.mol
This structure also has a valence error on the nitrogen. And really, if opened in 5.9.4, it shows the valence error, but when the nitrogen gets selected, then a + charge appears on it, and the valence error disappears.
I can only guess that the reason could be a conflict between Marvin's 'structure-to-name' feature, which "knows" that the correct structure of "(S)-cis-N-methylstylopine" should have a + charge on the nitrogen.
We are going to investigate this, and thank you for the reporting.

CPD-12945
From the link you have provided (http://metacyc.org/META/NEW-IMAGE?object=CPD-12945), I found that the molecule you cited was "Kinamycin D". Searching that on the net I found more reliable sources (which show the same molecular formula, and the same molecular structure) for this:
http://pubchem.ncbi.nlm.nih.gov/summary/summary.cgi?sid=582784&viewopt=PubChem />http://www.ebi.ac.uk/chebi/searchId.do?chebiId=CHEBI:31751 />Even the chemical formula of this molecule is different in your source! It is C23H18NO9 from your source, but C22H18N2O9 from the latter two.
So I would suggest you to be more careful with downloading molecules from less reliable sources. If you use the Chemical Abstracts (CAS) Registry Number (35303-14-1) or the Beilstein Registry Number (505073) you can identify the structure more reliably. Pubchem is also a quite reliable (free) source of information.

Regarding your comment "Posted: Fri Aug 31, 2012 6:20 pm", MSketch highlights the valence error always with a red underline, red circle is used probably by other software.

And again I would strongly suggest to be more careful with downloading molecules from the internet, as you can find a lot of conflicting (i.e. false) information on the internet about the same molecule.
So first please control your sources of information. (A similar case was in the https://www.chemaxon.com/forum/ftopic9082.html forum post as well.)

Best regards,

viktorw