SMARTS Substructure searches

In Excel,  I’m using =JCSubstructureMatchCount(J$2,$B31)
to look for specific substructures in SMILES structures.  

I have two specific questions. 

 

I have used a number of SMARTS like

CH (aromatic)	[cH1]
-C (aromatic)	[cH0]
N (aromatic)	n

Where the small letter c or n should mean carbon or nitrogen
in an aromatic context,  but the JCSubstructureMatchCount is finding them
in any context.  

Secondly, I would like to find, for example, a primary
alchohol, but not in a carboxylic acid.  ([Cv4][OH1]) finds primary
alcohols in any context.

 I would think that
  ([Cv4][OH1])!([CX3](=O)[OX2H1])  would work. But it doesn’t. 
Can you suggest a solution?

Hi James,

 

Regarding your second question, as a first idea we recommend to apply the following SMARTS expression:

[$([CX4]O)]

The following pages might also help you to create the appropriate SMARTS:

http://www.daylight.com/dayhtml/doc/theory/theory.smarts.html (see the Recursive SMARTS section)

http://www.daylight.com/dayhtml_tutorials/languages/smarts/smarts_examples.html

 

Best regards,

Krisztina

Hi James,

The issue is, that [n] is recognized as a simple N is SMILES and SMARTS. Please, use [n;a] or [#7;a] instead, see in attached example file.

You can check if your SMARTS is correct in MarvinSketch, if you draw the structure and select Edit/Source/View: SMARTS, SMARTS.

Let me know, if this helps.

Regards,

Anna

aforro wrote:

Hi James, The issue is, that [n] is recognized as a simple N is SMILES and SMARTS. Please, use [n;a] or [#7;a] instead, see in attached example file. You can check if your SMARTS is correct in MarvinSketch, if you draw the structure and select Edit/Source/View: SMARTS, SMARTS. Let me know, if this helps. Regards, Anna

Hi Anna, 

Thanks! 

This absolutely solved the problem!  I have referred to the Daylight documentation and I never saw the "a" functionality.  Is this a standard SMARTS nomenclature?

Thanks, Krisztina

That solved the problem. 

Can you explain what the function of $ is in this context?

Anna and Krisztina, 

As follow up questions. 

1. Similar to the alcohol but not carboxyllic acid, how do I specific and amine but not amide? I used [NX3;H2], but this also hits amides. 

2. How can I specify a primary secondary and tertiary alcohol eg CX3-OH, CX2H-OH, CXH2-OH?

Thanks

Alex 

Hi Alex,

I think the answers are found on the pages I linked in my previous post  -  I could not answer better.

On that page, please look for:

Recursive SMARTS

amine, not amide

If you need more detailed support we can assist within the frame of a consultancy project.

Best regards,

Krisztina