Clean2D on table standardizer

Dear reader,

For the molecule tables in our database we typically apply a number of standardizer operations to ensure that the molecule depiction in the database is consistent. One of the standardizer operations we use is the Clean2D option, which we use to show molecules from congeneric series in a similar fashion (using the template based alignment options in Standardizer).

The way I usually do this, is to use the standalone Standardizer, to generate a Standardizer xml file, which I then load in to my InstantJChem database (via the Entities > 'Molecule table' > Standardizer tab). 

Recently I wanted to make a change to one of the standardizer xml files I use. But I discovered that my InstantJChem version 6.2.0 will not read standardizer xml files which contain a Clean2D option with template based alignment if I generate them using Standardizer 6.x or higher. However if generate the same xml file (with Clean2D) using Standardizer 5.12.6,  there is no problem, and I can use this xml file to load into InstantJChem, and my molecules are nicely aligned based on the templates I provided.

Visual inspection of the 6.x and the 5.x Standardizer files reveal the format of the Clean2D template definition is very much different between the two files… And it looks that InstantJChem can only read the old definitions.

Is this expected behaviour?

Many thanks,

Sander.

I second to this that Clean 2D no longer seems to be an option one can choose in the Standardizer configuration panel of IJC 6.2.0.

Regards,

Evert

A word of caution here. Clean2D is not a recommended action for a database standardizer.

This is because it does not change the molecular graph in any way - its effect is entirely cosmetic. As such using it as database standardizer has no impact on search results, but may have an impact on search performance as unnecessary 2D cleaning will take place.

If you do want "cleaned" structures in the database the best approach is to clean them before importing them.

Tim

tdudgeon wrote:

This is because it does not change the molecular graph in any way - its effect is entirely cosmetic. As such using it as database standardizer has no impact on search results, but may have an impact on search performance as unnecessary 2D cleaning will take place. If you do want "cleaned" structures in the database the best approach is to clean them before importing them.

Dear Tim,

Thank you for your suggestion. 

The effect that I am after is indeed purely cosmetic. My goal (and also a strong request from our med chems) is to have congeneric series visualised in such a way in the database that they are all oriented in a similar fashion. And this also where part of the difficulty comes in: as SAR develops, certain R-groups might grow in size, requiring a different orientation of the scaffold to maintain a visually pleasing layout. This is impossible to predict beforehand, making it practically not feasible to only apply Clean2D only at import.

Applying a Clean2D standardizer to the database allows me to modify the templates at any point in time, keeping all molecules similarly aligned as SAR develops. 

Would it be technically possible to update molecules in a structure table with a new orientation? I looked at the possibility of importing an SDF file with updated orientations and then merging on the structure field, but this is not allowed by IJC.

Thanks,

Sander.

One possible approach could be a script that re-"standardized" the structures. Would read the structure, standardize it according to your needs and then write the structure back to the DB. When you needed to change your rules you just re-run the process.

The only risk of this is that you no longer have the original structure in the DB, but if all you are doing is Clean2D that may not be a worry.

Tim