Are these two structures tautomers?

A chemist thinks these two are tautomers and says jc_equal should return 1.

COC1=CC=C(C=C1)C2=CC(N=C3N=CNN23)C4=CC=C(Cl)C=C4

COC1=CC=C(C(C=C(C2=CC=C(C=C2)Cl)N3)N4C3=NC=N4)C=C1

Hi Yingyao,

He is probably right, the two structures seem tautomers.

However, jc_equals is not designed to recognize tautomers by default, it checks the hydrogens and the exact location of single/double bonds.

The following section of the documentation lists the available methods for handling tautomers:

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

 

Of these, option 1 (tautomer duplicate table or index option) is currently able to recognise the two as tautomers.

 

Our colleague working on the tautomer theory itself is on holiday at the moment. Next week he will be able to comment on the other methods.

Best regards,

Szabolcs

I try to enable tautomer in search (I am using 5.1.4), but it returns 0 (I also tried to aromatize both query and target, cannot get 1)?

SELECT jc_compare(
       'COC1=CC=C(C=C1)C2=CC(N=C3N=CNN23)C4=CC=C(Cl)C=C4',
       'COC1=CC=C(C(C=C(C2=CC=C(C=C2)Cl)N3)N4C3=NC=N4)C=C1',
't:e tautomer:y') FROM dual;

Hi Yingyao,

- Enabling the Tautomer search in jc_compare will enumerate some tautomers of the query according to some criteria. As Szabolcs mentioned our expert can aswer this part later.

- For duplicate search "perfect" search (in the latest version ranemed to "duplicate") is sugested, so t:p instead of t:e in your case.

- As Szabolcs noted the the preferred method of "Tautomer duplicate
filtering table option or
JChem
index option." already works for these structures, as it uses an other method called generic tautomers.

This is the recommended way of detecting tautomer duplicates.

This method does not depend on the number of possible enumerable tautomers.

In this case you do not have to specify the tautomer search: all PERFECT searches will aiutomatically detect the tautomers.

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

 

Best regards,

 

Szilard

 

 

 

1. I have tried both t:e and t:p, both return 0.

2. We cannot use the option 1, since we already have structures in the table that are tautomers. If we drop and rebuilt the index, I guess the indexing will fail according to document at http://www.chemaxon.com/jchem/doc/guide/cartridge/index.html#tautomer_duplicate_filtering

duplicateFiltering [y|n]: if set to "y", indexing and INSERT will check whether the structure being indexed are a duplicate of a structure already indexed in the same column. If the new structure is a duplicate, an exception is thrown.

I don't know if one can force the indexing by playing with haltOnError, never tried it before.

yzhou wrote:

1. I have tried both t:e and t:p, both return 0. 2. We cannot use the option 1, since we already have structures in the table that are tautomers. If we drop and rebuilt the index, I guess the indexing will fail according to document at http://www.chemaxon.com/jchem/doc/guide/cartridge/index.html#tautomer_duplicate_filtering duplicateFiltering [y|n]: if set to "y", indexing and INSERT will check whether the structure being indexed are a duplicate of a structure already indexed in the same column. If the new structure is a duplicate, an exception is thrown. I don't know if one can force the indexing by playing with haltOnError, never tried it before.

1: Please set the appropriate index option first (option 1)

2: You should set haltOnError to 'n'

http://www.chemaxon.com/jchem/doc/guide/cartridge/index.html#haltonerror

 

Szilard

There is a known issue in the current version of the cartridge:

Index parameter TDF=y works correctly only if duplicateFiltering=y is also set.

It will be corrected soon with version 5.2.1. From that version, TDF=y alone will also work correctly.

 

Sorry about the inconvenience.

 

Best regards,

Szabolcs

Hi,

A chemist thinks these two are tautomers and says jc_equal should return 1.

These two structures are not perfect tautomers of each others.   We don't consider carbon-carbon tautomerization step,blue line on the attached figure, this is why the two structures can not be converted to a common structure.

 

On the other hand a common tautomerization path exist. Red line on the attached figure.

 

Jozsi

Thanks for the explanation, if you don't consider them as tautomers, why current jchem recognizes them as tautomers?

1. The post above suggests setting "(tautomer duplicate table or index option)" is able to treat the two structures as replicates.

2. We tried to use Standardizer to general cannonical tautomers, which does produce the same smiles from both input, it converts

COC1=CC=C(C=C1)C2=CC(N=C3N=CNN23)C4=CC=C(Cl)C=C4  

into 

COC1=CC=C(C(C=C(C2=CC=C(C=C2)Cl)N3)N4C3=NC=N4)C=C1

Hi,

...why current jchem recognizes them as tautomers?

This is a bug which is related with the above mentioned 'carbon-carbon' tautomerization step.

We have fixed this bug. I hope the fixed version will appear in the next patch release.

 

Jozsi

Jozsi wrote:

Hi, ...why current jchem recognizes them as tautomers? This is a bug which is related with the above mentioned 'carbon-carbon' tautomerization step. We have fixed this bug. I hope the fixed version will appear in the next patch release.   Jozsi

We just upgraded to v5.2.3.1 turning on TDF=y option, and tried the following query, but it seems that tautomer search is still broken since it returns to 0, not 1.  Could you check on your side?

SELECT jc_compare(
       'COC1=CC=C(C=C1)C2=CC(N=C3N=CNN23)C4=CC=C(Cl)C=C4',
       'COC1=CC=C(C(C=C(C2=CC=C(C=C2)Cl)N3)N4C3=NC=N4)C=C1','t:d tautomer:y') FROM dual

Hi,

 

These molecules are not tautomers of each other. The "return value 0"  is correct. 

See my note  in this topic above.

 

 

 

 

Jozsi

Jozsi wrote:

Hi,   These molecules are not tautomers of each other. The "return value 0"  is correct.  See my note  in this topic above.         Jozsi

Hi Jozsi

Sorry, I was not careful to read all comments.  Anyway, how about this case? I confirmed that jchem does not recognize the following two structures are tautomer, but is it correct? If you cofirm the reasoning behind, I will appreciate.

Thanks,

yasushi

 

 

Hi Yasushi,

These two structures are tautomers of each others,of course. (green rectangle on the figure, I have attached). Their generic tautomers are the same. You can check this fact  even with Marvin's  trial version at our home page.  It means that JChem should recognize these two structures in the green rectangle as identical ones.

 

I can't image why you impacted with this bug. We check in JChem and fix it ASAP!!

 

 

Jozsi

 

 

Jozsi wrote:

Hi Yasushi, These two structures are tautomers of each others,of course. (green rectangle on the figure, I have attached). Their generic tautomers are the same. You can check this fact  even with Marvin's  trial version at our home page.  It means that JChem should recognize these two structures in the green rectangle as identical ones.   I can't image why you impacted with this bug. We check in JChem and fix it ASAP!!     Jozsi

Hi Jozsi

I am not quite sure if I understood correctly what you are saying.  Two structures I draw was marked in red in your post and tagged as non tautomer couple.  But it sounded you agreed they were tatumer of each others.

We are testing v5.3.2.1.  Pleast check if the tautomer search returns with the other stucture. It is not returning in our test.

 

Thanks,

yasushi

 

Dear Yasushi-San!

 

Józsi in his post on Posted: 2009 07 17, Fri 18:42  showed which two structures are not tautomers and which are.

Now I checked their matching in search:

 jcsearch --tautomer:y -q 'C1C=CNC2=NC=NN12' 'C1C=CN=C2N=CNN12'
C1C=CN=C2N=CNN12

This means the green ones are tautomers. They match in perfect search as well:

wagner@rwagner:~/work/jchem$ jcsearch -t:p --tautomer:y -q 'C1C=CNC2=NC=NN12' 'C1C=CN=C2N=CNN12'
C1C=CN=C2N=CNN12

The red ones are the followings:

wagner@rwagner:~/work/jchem$ jcsearch -t:p --tautomer:y -q 'C1C=CNC2=NC=NN12' 'C1C=CN2NC=NC2=N1'
wagner@rwagner:~/work/jchem$ jcsearch -t:s --tautomer:y -q 'C1C=CNC2=NC=NN12' 'C1C=CN2NC=NC2=N1'

They are not matching, because they are not tautomers.

Now I try to clarify earlier posts:

In your post on Posted: 2009 07 17, Fri 17:11  you were complaining that the green ones are not matching. For us they match, do they not match for you?

In your post on Posted: 2009 07 17, Fri 15:26  you were also telling that two structures do not match, they don't match for us either. Those are the structures which were drawn red and shouldn't be recognized as tautomers.

Hope this helps.

Robert

rwagner wrote:

Dear Yasushi-San!   Józsi in his post on Posted: 2009 07 17, Fri 18:42  showed which two structures are not tautomers and which are. Now I checked their matching in search:  jcsearch --tautomer:y -q 'C1C=CNC2=NC=NN12' 'C1C=CN=C2N=CNN12' C1C=CN=C2N=CNN12 This means the green ones are tautomers. They match in perfect search as well: wagner@rwagner:~/work/jchem$ jcsearch -t:p --tautomer:y -q 'C1C=CNC2=NC=NN12' 'C1C=CN=C2N=CNN12' C1C=CN=C2N=CNN12 The red ones are the followings: wagner@rwagner:~/work/jchem$ jcsearch -t:p --tautomer:y -q 'C1C=CNC2=NC=NN12' 'C1C=CN2NC=NC2=N1' wagner@rwagner:~/work/jchem$ jcsearch -t:s --tautomer:y -q 'C1C=CNC2=NC=NN12' 'C1C=CN2NC=NC2=N1' They are not matching, because they are not tautomers. Now I try to clarify earlier posts: In your post on Posted: 2009 07 17, Fri 17:11  you were complaining that the green ones are not matching. For us they match, do they not match for you? In your post on Posted: 2009 07 17, Fri 15:26  you were also telling that two structures do not match, they don't match for us either. Those are the structures which were drawn red and shouldn't be recognized as tautomers. Hope this helps. Robert

Hi Robert

Thanks for your clarification.  I will check with our chemist.  The two structures I posted were given by our chemist.  He may misdrawn a structure.

 

Thakns,