Unusual result from jc_insert function

I have inserted a row into the database using the jc_insert operator. The code I used is as below:

Code:

aCDReturn := jchem_table_pkg.jc_insert(vSmiles, 'jchemistry', null, 'true', 'false', ''); iCDNum := aCDReturn(1);

The string for vSmiles is calculated from Accord for Excel and then sent to the PL/SQL procedure. This works fine. However I have come across a problem to do with aromaticity.





Structure 1 is attached and was registered into the database. The SMILES string calculated by Accord for Excel was c1ccc2c(c1)C(\C=C/O2)=O





However when this is written to the database in the PL/SQL procedure the CD_SMILES column in the JCHEMISTRY table shows O=c1ccoc2ccccc12 as the SMILES string. Notice the lower case atoms, that signify aromaticity. I may be wrong, but I did not think my original molecule was aromatic, yet the cartridge has attempted to draw it as aromatic.





When I produce an image of the inserted structure the cartridge "draws" the pyran ring as aromatic - see Structure 2.





Surely this is incorrect?

ftm wrote:

Surely this is incorrect?

It is right, the structure is aromatic by the default "General" aromatization.


You may prefer the other aromaticity option: "Basic".





See the documentation about these aromatization options:


http://www.chemaxon.com/marvin/doc/user/aromatization-doc.html





You may also find the following discussion interesting:


http://www.chemaxon.com/forum/ftopic2530.html


(particularly posts after Wed Jan 31, 2007)





It is possible to use basic aromatization in JChem Base and Cartridge through custom standardization:


http://www.chemaxon.com/jchem/doc/user/Query.html#standardizationDB


http://www.chemaxon.com/jchem/doc/user/StandardizerConfiguration.html#aromatizesec





Let us know if you have more questions.





Szabolcs

Just one more addition: JChem always uses the cd_smiles column for searching. However, you can use the cd_structure column for visualization. This always contains the original source of the molecule.

Hi Szabolcs





Thanks for the replies. Is there an easy way of setting the aromaticity flag to 'Basic' as an option when using the JC_MOLCONVERTB operator or is this something that has to be done with the Standardizer?





I was interested to read your reply about using the CD_STRUCTURE column for reporting structures. When I try to use this in my PL/SQL procedure it gives an error of





Error: PLS-00306: wrong number or types of arguments in call to 'JC_MOLCONVERTB'





I used the code as below. The original select clause using the CD_SMILES column is commented out. It works on CD_SMILES but errors with CD_STRUCTURE. I'm guessing that I will have to convert the BLOB returned into a format suitable for the JC_MOLCONVERTB operator prior to creating the image. Perhaps a MOL text? Is this correct?

Code:

create or replace procedure chemblob_to_file(myfilename in varchar2, chemistry in number) is v_blob         blob; blob_length    integer; out_file       utl_file.file_type; v_buffer       raw(32767); chunk_size     binary_integer := 32767; blob_position  integer := 1; begin -- Retrieve the BLOB --select jc_molconvertb(cd_smiles, 'jpeg') select jc_molconvertb(cd_structure, 'jpeg') into v_blob from jchemistry where cd_id = chemistry; -- Calculate size of BLOB blob_length := dbms_lob.getlength(v_blob); -- Open handle to write file location -- wb parameter is "write in byte mode" out_file := utl_file.fopen('NEW_DIR', myfilename, 'wb', chunk_size); -- Write the BLOB to file in chunks while blob_position <= blob_length loop       if blob_position + chunk_size - 1 > blob_length then          chunk_size := blob_length - blob_position + 1;       end if;       dbms_lob.read(v_blob, chunk_size, blob_position, v_buffer);       utl_file.put_raw(out_file, v_buffer, true);       blob_position := blob_position + chunk_size; end loop; -- Close the file handle utl_file.fclose(out_file);   end chemblob_to_file;

Quote:

Is there an easy way of setting the aromaticity flag to 'Basic' as an option when using the JC_MOLCONVERTB operator

Please use the smiles:a_bas option with jc_molconvert (http://www.chemaxon.com/marvin/doc/user/molconvert.html):

Code:

select jc_molconvert('O=C1C=COc2ccccc12', 'smiles:a_bas') from dual

You can also use jc_standardize:

Code:

select jc_standardize('O=C1C=COc2ccccc12', 'config:aromatize:b') from dual;

If you need this standardization for all your structures in the table, I suggest to consider setting a custom standardization for the table (in the case of JChem structure tables) or the jc_idxtype index (in the case of regular structure tables). The way to specify this with regular structure tables is through an index parameter:

Code:

CREATE INDEX jc_idx_2 ON jchemtest2(structure_col) INDEXTYPE IS jc_idxtype PARAMETERS('std_config=aromatize:b')

. With JChem structure tables you have to specify the custom standardization configuration while creating the table with JChem Manager (http://www.chemaxon.com/jchem/doc/admin/index.html#create). Note that JChem Manager accepts only Standardizer configuration files in XML format. The XML format of the basic aromatization configuration looks like this (http://www.chemaxon.com/jchem/doc/user/StandardizerConfiguration.html#config,


http://www.chemaxon.com/jchem/doc/user/StandardizerConfiguration.html#aromatizesec):

Code:

<?xml version="1.0" encoding="UTF-8"?> <!-- Standardizer configuration file --> <StandardizerConfiguration Version ="0.1">     <Actions>       <Aromatize ID="chemaxonaromatize" Type="basic"/>     </Actions> </StandardizerConfiguration>

When your structure table (index) has a custom standardization configuration assigned to it, the standardization will be done automatically for you.

Quote:

wrong number or types of arguments in call to 'JC_MOLCONVERTB'

With cd_structure, please, use jc_molconvertbb, which accepts BLOB as its first parameter. (Note the double 'b'.)





Peter

Many thanks for the response. Apologies, but I still cannot get this to work for the jc_molconvertb function. I require a BLOB of the image to be created and I then save this to the local disk.





I have tried:

Code:

select jc_molconvertb(jc_standardize(cd_smiles, 'config:aromatize:b'), 'jpeg') into v_blob from jchemistry where cd_id = chemistry;

and

Code:

-- Retrieve the BLOB select jc_molconvertb(cd_smiles, 'jpeg:a_bas') into v_blob from jchemistry where cd_id = chemistry;

Neither are returning an error but the image files (JPG) that are produced are still in the GENERAL aromatic form (see attached).





Also, when I try to use jc_molconvertbb on my CD_STRUCTURE column I get an error (also attached as error.jpg) Does this mean I can only index either the SMILES column or the STRUCTURE column? If so then I can't return an image from the contents of the STRUCTURE column.

Quote:

the image files (JPG) that are produced are still in the GENERAL aromatic form

Is it not that the cd_smiles column already contains the aromatized form? (I think the aromatize action will never ever dearomatize an already aromatized ring.)

Quote:

when I try to use jc_molconvertbb on my CD_STRUCTURE column I get an error (also attached as error.jpg). Does this mean I can only index either the SMILES column or the STRUCTURE column?

You should be able to create a jc_idxtype index on both columns. But for your immediate purpose, I'd suggest to use the jcf_molconvertbb function instead of the jc_molconvertbb operator (note the 'f' in jcf_). The function will not complain about indexes and will not be worse than the operator performance-wise either.





Peter