MarvinSketch 5.0 repeated file corruption

I've lost hours of work in the last day as Marvin Sketch 5.0 seems to be repeatedly corrupting the files I'm working with. I'll save over a previous version of the file, and don't get any error message. But when I open Marvin Sketch later and try to re-open the file it cannot be read. The 5.0alpha seemed more stable - I never had this happen!





I'm running on Mac OS 10.4.11 on a G4, and set up a Markush structure for the enumeration plugin.





The error dialog says:





"Cannot read molecule 1"





I can't seem to get the text of the stack trace, so a screen cap is attached

What's particularly frustrating is that it seems that this corruption seems to happen while using MarvinSketch and then there is no way to recover - doing save-as to a new file name does not produce a usable file.





Also, copy/paste within MarvinSketch stops working, suggesting that it's the internal representation that is broken in some way.





There is no error message or other obvious indication when this has occurred.

Please send a corrupted file to check it.

the corrupted files have some confidential information - I can't post publicly. I'm trying to see it happen with a new file.





the starting file that all these corrupted ones are derived from was created using the 5.0alpha - is there any chance that's the cause?

I will contact you in email regarding this issue.

Also, when trying to ennumate the Markush structures, I was seeing random additional, unattached, methyl groups showing up. I seemed to be able to get them to go away if I re-selected a set of possible substituants (e.g. R3) and hitting R3 again.





I'm not sure if this could be related to the invalid/corrupted structures in the file.

I have checked your MRV file (that you sent it in private), it includes also two methanes (you can find them among the fragments of R3). So they are not extra atoms in the result of enumeration.


Please, clean the original structure to render fragments. After then, you can see that methanes are there.


Probably, you placed them to the wrong place or missed to add them to the R-group definition.