molconvert for haem units

Hi,





I was wondering if I could get some help: I'm using molcovert in version 3, though as a test I also tried the molconvert in 4.0 from the web site and it seems to manifest the same problem.





Basically - molconvert seems to have real trouble making decent 2D images of HAEM units: It seems that quite often, the converter mangles up one of the rings.





I've tested this on a lot of HAEM units from the PDB and I'd say nearly all of them get corrupted in this way. It also doesn't seem to matter what bond typing convention I adopt.





I've tried various optimisation options (dont optimise, always optimise) but to no avail. I also have tried first reducing an SD file smiles string using molconvert, then rebuilding as a 3D SD file and then passing this in to the image conversion (oddly, this seems to work better for some molecules than just going straight from a third party SD file to an image. Guess it is better at getting the molecule into molcovert's own conventions)





Has anyone seen anything similar, and resolved it?

Hi,





Thank you for the bug report.


We will improve the 2D cleaning in the near future. This means that the whole code will be rewritten. This is neccessary for some reason, but we have also noticed that the 2D representation of some other molecule is also bad.


About the schedule:


The code revision starts within few weeks.


The first beta version of the new code probably out in few month, let's say around July. I am sorry for not giving exact date, but that's all I know right now.








Andras

Hi Andras,





I have several other examples (not HAEMS) that manifest similar problems - if you'd like a SD file containing these then let me know, and I'll post it.





Jason

Hi,





I would be glad if you would attache the SDfile for testing purposes.





Thank you


Andras