isothiocyanates cannot be cleaned?

Another one - Clean3D (fine setup) breaks down with isothiocyanates (i suppose, that's the most exotic group!) but Clean3D (fast) seems to do the job more or less well. Molecule was:





CC(=O)Nc1:c:c:c(\C=C\c2:c:c:c(:c:c2S([O-])(=O)=O)N=C=S):c(:c1)S([O-])(=O)=O





The funny thing is that stupid plain phenylisothiocyanate works fine - so maybe it ain't the -NCS group, after all?

Hi Dragos,





What do you mean in "breaks down"? The result seems to be acceptable for me and cleaning with "fine" in both, MSketch and molconvert works. I asked the guys around to try it, if it is version or operating specific.





Best wishes,





Ödön

dragos wrote:

Another one - Clean3D (fine setup) breaks down with isothiocyanates (i suppose, that's the most exotic group!) but Clean3D (fast) seems to do the job more or less well. Molecule was: CC(=O)Nc1:c:c:c(\C=C\c2:c:c:c(:c:c2S([O-])(=O)=O)N=C=S):c(:c1)S([O-])(=O)=O The funny thing is that stupid plain phenylisothiocyanate works fine - so maybe it ain't the -NCS group, after all?

In my hands, it just says "Clean Failed" and then gets a NullPointerException when trying to access the conformer list. In MSketck I get a Clean Failed as well - however, it is true that eventually the conformers that are displayed are correct. I have made changes in the Conformers code in order to catch this errors (and go on with a single geometry when such bizzare things happen). The code in its present form successfully built 2300 drug & druglike compounds in a matter of several hours (say one night) on x86_64 - and, when you process big molecule sets you prefer to have a few bad or unvalidated conformers than crashes of your halfways done job. So I'm reasonably happy for the moment ;-)

Hi Dragos,





Please, give me the exact version of the used Marvin, because none of us could reproduce the error.





Best wishes,





Odon

dragos wrote:

In my hands, it just says "Clean Failed" and then gets a NullPointerException when trying to access the conformer list. In MSketck I get a Clean Failed as well - however, it is true that eventually the conformers that are displayed are correct. I have made changes in the Conformers code in order to catch this errors (and go on with a single geometry when such bizzare things happen). The code in its present form successfully built 2300 drug & druglike compounds in a matter of several hours (say one night) on x86_64 - and, when you process big molecule sets you prefer to have a few bad or unvalidated conformers than crashes of your halfways done job. So I'm reasonably happy for the moment ;-)

Dragos,


When you press the "New topic" or "Reply" button, you see this message on the left:


"When posting support requests please always specify the Marvin or JChem version in use, and also software and hardware environment and versions (when relevant)."


It would save us a lot of time if people took that seriously :-)

It's marvin 4.0.4 - but, oops, I checked it on the 32-bit laptop running Suse 9.1 - and it worked well! It only crashes on the new 64-bit HP machines running Suse 10! So much for the progress! Sorry - could not have guessed that good old Euclidean 3D stuff may depend on versions and hardware (?)

Just to make it more interesting, my HP nx6125 laptop is also 64-bit, running SuSE 10. It is probably the java version which causes the problem.


At my machine:


$ java -version


java version "1.4.2_10"


Java(TM) 2 Runtime Environment, Standard Edition (build 1.4.2_10-b03)


Java HotSpot(TM) Client VM (build 1.4.2_10-b03, mixed mode)





Best wishes,





Odon

dragos wrote:

It's marvin 4.0.4 - but, oops, I checked it on the 32-bit laptop running Suse 9.1 - and it worked well! It only crashes on the new 64-bit HP machines running Suse 10! So much for the progress! Sorry - could not have guessed that good old Euclidean 3D stuff may depend on versions and hardware (?)

Yep - maybe! I'm on





java version "1.5.0_03"


Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0_03-b07)


Java HotSpot(TM) 64-Bit Server VM (build 1.5.0_03-b07, mixed mode)