update jchem tables

User 6b1e802ce9

28-04-2011 14:17:25

Every time we upgrade a JChem version, the tables need to regenerate.  A very time consuming process, which is slower on every newer version. 


For a small 10K library it is taking 3 days.  We have another library with over 30 Million structures.


Do you know of any alternatives or options to reduce the time?  Besides dropping the table and import - again.

ChemAxon a3d59b832c

29-04-2011 07:37:31

Hi Caty,


 


Regeneration should not take that long. We are aware that in JChem version 5.4.0 table regeneration for tautomer duplicate tables was very slow, but that issue has been fixed in the most recent version. (5.4.1)


For 10K structures it should not take more than a few minutes.


 


Which JChem version are you using, and do you use any extra table settings?


(Like tautomer duplicate option, calculated columns, structural keys, standardization, etc.?) - These might effect regeneration time, but I do not expect that large slowdown in any case.


 


Best regards,


Szabolcs

User 6b1e802ce9

04-05-2011 18:22:09

The version upgrade was from 5.3.6 to 5.4.1.1


We added 252 structural keys. 


It seems that the machine also crashed almost at the end.  I dropped and created a new table, added structural key and the import is quite slow too.


Can structural keys slow down?

User 6b1e802ce9

04-05-2011 18:31:26

17 new molecules, it took 1052 seconds (~17 min)

ChemAxon 9c0afc9aaf

04-05-2011 22:10:12

Hi,


Can structural keys slow down?

Yes, structural keys requre additional computations during import and table recalculation, slowing these opreations down. The slowdown may depend on the number of keys, the keys used and the imported data.


http://www.chemaxon.com/jchem/doc/dev/dbconcepts/index.html#structuralKeys


Even with the high number of structural keys the 1 molecule per second you have reported sounds excessively slow.


If you send us the key and the 17 structures we can take a look.


Best,


Szilard