Marvin chokes on loading some cyclic molecules

I am using Marvin Beans/MarvinView 3.4.3. For certain mol files, when I try to load them in (and nothing more) it seems to take forever, using 100% CPU (I let it go about an hour before killing it).





An example is DEGLUCOBALHIMYCIN from PDB 1HHY (converted to .mol format). I attached the exact MOL file I am using if that helps.

finesse wrote:

I am using Marvin Beans/MarvinView 3.4.3. For certain mol files, when I try to load them in (and nothing more) it seems to take forever, using 100% CPU (I let it go about an hour before killing it). An example is DEGLUCOBALHIMYCIN from PDB 1HHY (converted to .mol format). I attached the exact MOL file I am using if that helps.

I suggest you download Marvin 3.5. I was able to load your file into Marvin 3.5 without any trouble. Visit http://www.chemaxon.com/test/marvin/doc/dev/example-sketch1.1.html to try it out.

The problem is probably that you have set chiralitysupport = All or R/S labels = all.


Please use chiralitysupport=none or set the


View-> Misc ->R/S labels -> none





We will speed up the Chirality calculation in the near future.





All the best


Andras

Yes that is it. If I turn off chirality it loads fine. If I try to set it to all, or to show SMILES, then it goes off into lala land. I hope you can fix this as calculating SMILES and R/S is the main thing we are using MarvinBeans for! It is a very handy tool, glad you made it!

It only happens when there are many chirality centers. Some kind of combinatorial explosion.

Actually, that is not it at all, I have figured it out and ti appears to be a 'bug' in Marvin. The 'bad' file I posted has two bonds which are exactly the same, repeated in the bond list (63 - 3). This apparently causes Marvin to go into and endless loop. This sounds like something that's easy enough to fix, so hopefully you can incorporate a fix into the next version :)





Note the same problem occurs if the bond order is reversed (ie if you have one bond between x-y and another between y-x)

Thanks for the bug report.


The bugfix will be available in the next Marvin release (3.5.1).