Combing smarts into a complex expression

Hello,

I would like to express the smarts for a class of compounds with an aliphatic carbon chain of 13 to 21 carbon atoms, where  the longest chain of carbon has between 13 and 21 carbons and is attached to a Nitrogen atom. Examples include:

    - CCCCCCCCCCCCCNH2 ; CCCCCC(C)C\C=C\C(C)CCCNH2 ;  CCCCCCCCCCC\C=C\CCCCNH2

 However, I do not want to include compounds containing a carbon chain of more than 21 carbon atoms. Which means that the following compounds should not be included:

   -  CCCCCCCCCCCCCCCC\C=C\CCCCNH2; CC/C=C\CCCCCCCCCCCC\C=C\CCCCNH2

 

I first generated the patterns for each of the two types:

13 carbons attached to N: (A)  =>  [#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-[#7]([H])[H]

22 carbons attached to N: (B)  =>  [#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-[#7]([H])[H]

I then tried to combined those into  1 smarts [$(A)!$(B)] * to mean that the compound of interest should match A but not match B. This does not work! 

What would be the best way of describing this in one SMARTS string?

[$(A)!$(B)]: [$([#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-[#7]([H])[H])!$( [#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-,=[#6]-[#7]([H])[H])]

Thank you

Regards,

MrYan

Hi Yan,

I've moved your topic to a more relevant place, my colleagues will answer soon.

 

Tamas Fazekas-Vigh

Hi MrYan,

 

I am not sure whether we interpret your question right, but here is a more simple query that will hit all the 13-21 "C" atom containing aliphatic amines:

In cxsmarts format:

[#6;A]~[#6;A][#7] |Sg:n:1:12-20:ht|  having any aliphatic bonds

or

[#6;A][#6;A][#7] |Sg:n:1:12-20:ht|   having only single aliphatic bonds

 

Best regards,

Norbert

Hello Norbert,

Your interpretation of my question is right. Thank you for your feedback.

Is it possible to have this expressed as a daylight smarts expression?

Regards,

MrYan

Hi,

I'm not sure, but maybe with recursive SMARTS you can describe your structure.

I have never tried this, but the following page could be a good starting point. Also there are examples here which could be examined to create your specific SMARTS expression.

I would try to enumerate the carbon chains in the recursive SMARTS part, that may work.

Hi Volfi,

I have visited the page already. Thank you.

Regards,

MrYan