Tautomerize fails to produce a single canonical tautomer

User 81a38f9467

12-11-2010 04:22:31

I am running Standardizer 5.3.8.  When I input more than one tautomer of a compound and use the Tautomerize function, the output is not always a single canonical tautomer.  Shouldn't the Tautomerize function produce one and only one canonical tautomer for each molecule? 


Here is the first example: Four tautomers of a pyrazolopyrimidinone are input;  three tautomers give the same canonical tautomer, but the fourth gives a different result.

User 81a38f9467

12-11-2010 04:31:20

Two tautomers of losartan give different results after using Tautomerize in the Standardizer:

User 851ac690a0

12-11-2010 08:10:04

Hi,


These bugs were fixed.  The fixed  version  "5.4" will be available in a couple of day.


Jozsi 

User 81a38f9467

12-11-2010 19:25:01

Here is a third example.

User 81a38f9467

12-11-2010 19:28:24

Hi Jozsi,


Thank you for your quick response.  I am looking forward to the new release (5.4) of the Standardizer and hope that these problems will be resolved.


Tim

User f914e5718c

12-11-2010 21:24:13

Jozsi,


The imidazole case was posted after you responded.  Just want to make sure the new release will solve all of these problems.


How long has this software been in production?  It seems like these flaws are with reasonably common structures


Can you give us a better time frame on the release? 


Does this mean we have to upgrade the whole cartridge again?  That was quite a lot of effort.


Brock.

User 851ac690a0

17-11-2010 13:20:33

Hi,


Yes, the new release will solve all of these problems.


For the rest of your questions I quote one of my collegaue:


"We understand that upgrading can be inconvenient. Tautomerism is a
complex problem of chemistry, and we do our best to provide a
reliable and high performance tautomer prediction tool. However, it
happens sometimes, that we find cases that are not handled properly.
By the nature of the problem, we cannot guarantee a perfect
prediction, but we extend our test base continuously, run tests
every night, and fix issues immediately."


 


Jozsi

User f914e5718c

17-11-2010 17:20:53

 


Yes -- I was being harsh.  I guess I was surprised by the relative common cases not working (e.g. imidazole).  I am trying to get a handle on how "mature" this software is.


The big question for us is when is the release date for this version?


Can we get a beta and try it out?  


The other big question is when is the next release date (after this one) -- should we just wait?


 

ChemAxon d76e6e95eb

17-11-2010 18:09:35

You were right, not harsh :)


You can get a release candidate in this week. We will be pleased if you give quick feedback before the actual release.

User 851ac690a0

17-11-2010 18:41:08

Hi,


Yes, I  agree. The "big question" part is right.


Jozsi