Tautomer search performance

Hello





We have noticed that the tautomer search is quite slow for some macrocycles.


For other compound types, we hardly notice the difference.


We are using JChem5.1.1


How easy would it be to improve the performance?





Thanks


Catherine





PS This is my example and the time it takes with Tautomer off / on.





select * from structures


where jc_compare(displayed_smiles,'CC(C)(C)OC(=O)N[C@H]1CCCCC\C=C/[C@@H]2C[C@]2(CC(=C)[C@@H]2C[C@H](CN2C1=O)OC(=O)N1Cc2cccc(F)c2C1)C(=O)NS(=O)(=O)C1CC1 |r,c:14|','t:ef exactIsotopeMatchingOption:i exactRadicalMatchingOption:i exactChargeMatchingOption:i doubleBondStereo:A exactStereoMatching:N stereoSearch:N tautomer:n')= 1


--422msec








select * from structures


where jc_compare(displayed_smiles,'CC(C)(C)OC(=O)N[C@H]1CCCCC\C=C/[C@@H]2C[C@]2(CC(=C)[C@@H]2C[C@H](CN2C1=O)OC(=O)N1Cc2cccc(F)c2C1)C(=O)NS(=O)(=O)C1CC1 |r,c:14|','t:ef exactIsotopeMatchingOption:i exactRadicalMatchingOption:i exactChargeMatchingOption:i doubleBondStereo:A exactStereoMatching:N stereoSearch:N tautomer:y')= 1


-- 1min32

Hi Catherine,





Indeed, molecules with many tautomers are slow to search with the tautomer option. There is an alternative, more efficient approach that involves the tautomer duplicate table/index parameter. See more details here:





http://www.chemaxon.com/jchem/doc/guide/dbconcepts/index.html#tautomers





Best regards,


Szabolcs