joelib-devel

[Joelib-devel] smarts

[naji h. <sun...@or...>]

HI all,

By using Smarts patterns I  don't manage to differ some molecu=
lar fragments (see below).
It's a problem because each group has a  diff=
erent contribution in order to calculate a property.

For example: I do=
n't know how I could make (with Smarts patterns) to differ those groups.=0D
=
>C=3D (participating in two fused benzene rings)
>C=3D (participating in=
 two fused non-benzene rings)
>C=3D (participating in two fused  rings: =
one benzene ring and one non-benzene ring)
-CH=3D( participating in a be=
nzene ring)
>C=3D (the formal double bond and a formal single bond parti=
cipating in a benzene ring)
>C=3D (two single bonds participating in a n=
on- benzene ring)
-CH=3D( participating in a non- benzene ring)
>C=3D (=
the formal double bond and a formal single bond participating in a non-be=
nzene ring)
=3DN- (participating in a ring )
>NH(participating in a rin=
g )
-N< (participating in a ring )
-N< (participating in two fused ring=
s )

Any suggestions?

Thanks

Naji


 

Re: [Joelib-devel] smarts

[Joerg K. W. <we...@in...>]

Hi naji,

so here are my guesses, no gurantee that the SMART patterns work. So 
please resend the tested/working patterns.

>>C= (participating in two fused benzene rings)
c~[#6]~c two benzene rings connected via aliphatic/aromatic via any bonds
or general aromatic rings ?
[r;a]~[#6]~[r;a]

>>C= (participating in two fused non-benzene rings)
[r;!a]~[#6]~[r;!a] two non-aromatic rings connected via 
aliphatic/aromatic via any bonds

>>C= (participating in two fused  rings: one benzene ring and one non-benzene ring)
[r;!a]~[#6]~[r;a]

> -CH=( participating in a benzene ring)
> 
>>C= (the formal double bond and a formal single bond participating in a benzene ring)
[#6,a,Q2,$(*-*),$(*=*)] so two heavy weight neighbors and neighbour 
environments with $(...), here general aromatic ring.
You can also work with hybridisation states using ^2, see 
joelib/data/plain/*.txt for examples

>>C= (two single bonds participating in a non- benzene ring)
with or without implicite hydrogens, so X or D or Q
http://www-ra.informatik.uni-tuebingen.de/software/joelib/tutorial/functionalities/mol-smarts.html
and reference and link to daylight tutorial.
[#6,!a,r,Q2,$(*-*),$(*-*)] so two heavy weight neighbours, here general 
non-aromatic ring

> -CH=( participating in a non- benzene ring)
[#6,!a,r,Q2,$(*-*),$(*=*)]

>>C= (the formal double bond and a formal single bond participating in a non-benzene ring)
difference to -CH= ?

> =N- (participating in a ring )
[#7,r,Q2,$(*-*),$(*-*)] so assuming any ring, or only aliphatic ones ?

>>NH(participating in a ring )
[#7,r,H1,$(*-*),$(*-*)]

> -N< (participating in a ring )
> -N< (participating in two fused rings )
analogue to C, more or less.
> Any suggestions?
Yes :-)

> Thanks
Was me a pleasure.

Kind regards, Joerg
-- 
Dipl. Chem. Joerg K. Wegner
Center of Bioinformatics Tuebingen (ZBIT)
Department of Computer Architecture
Univ. Tuebingen, Sand 1, D-72076 Tuebingen, Germany
Phone: (+49/0) 7071 29 78970
Fax: (+49/0) 7071 29 5091
E-Mail: mailto:we...@in...
WWW:    http://www-ra.informatik.uni-tuebingen.de
--
Never mistake motion for action.
                                     (E. Hemingway)

Never mistake action for meaningful action.
                                (Hugo Kubinyi,2004)

Re: [Joelib-devel] smarts

[Joerg K. W. <we...@in...>]

Hi,

sorry, replace all
,=or statement
;=and statement

> Hi naji,
> 
> so here are my guesses, no gurantee that the SMART patterns work. So 
> please resend the tested/working patterns.
> 
>>> C= (participating in two fused benzene rings)
> 
> c~[#6]~c two benzene rings connected via aliphatic/aromatic via any bonds
> or general aromatic rings ?
> [r;a]~[#6]~[r;a]
> 
>>> C= (participating in two fused non-benzene rings)
> 
> [r;!a]~[#6]~[r;!a] two non-aromatic rings connected via 
> aliphatic/aromatic via any bonds
> 
>>> C= (participating in two fused  rings: one benzene ring and one 
>>> non-benzene ring)
> 
> [r;!a]~[#6]~[r;a]
> 
>> -CH=( participating in a benzene ring)
>>
>>> C= (the formal double bond and a formal single bond participating in 
>>> a benzene ring)
> 
> [#6,a,Q2,$(*-*),$(*=*)] so two heavy weight neighbors and neighbour 
> environments with $(...), here general aromatic ring.
> You can also work with hybridisation states using ^2, see 
> joelib/data/plain/*.txt for examples
> 
>>> C= (two single bonds participating in a non- benzene ring)
> 
> with or without implicite hydrogens, so X or D or Q
> http://www-ra.informatik.uni-tuebingen.de/software/joelib/tutorial/functionalities/mol-smarts.html 
> 
> and reference and link to daylight tutorial.
> [#6,!a,r,Q2,$(*-*),$(*-*)] so two heavy weight neighbours, here general 
> non-aromatic ring
> 
>> -CH=( participating in a non- benzene ring)
> 
> [#6,!a,r,Q2,$(*-*),$(*=*)]
> 
>>> C= (the formal double bond and a formal single bond participating in 
>>> a non-benzene ring)
> 
> difference to -CH= ?
> 
>> =N- (participating in a ring )
> 
> [#7,r,Q2,$(*-*),$(*-*)] so assuming any ring, or only aliphatic ones ?
> 
>>> NH(participating in a ring )
> 
> [#7,r,H1,$(*-*),$(*-*)]
> 
>> -N< (participating in a ring )
>> -N< (participating in two fused rings )
> 
> analogue to C, more or less.
> 
>> Any suggestions?
> 
> Yes :-)
> 
>> Thanks
> 
> Was me a pleasure.
> 
> Kind regards, Joerg


-- 
Dipl. Chem. Joerg K. Wegner
Center of Bioinformatics Tuebingen (ZBIT)
Department of Computer Architecture
Univ. Tuebingen, Sand 1, D-72076 Tuebingen, Germany
Phone: (+49/0) 7071 29 78970
Fax: (+49/0) 7071 29 5091
E-Mail: mailto:we...@in...
WWW:    http://www-ra.informatik.uni-tuebingen.de
--
Never mistake motion for action.
                                     (E. Hemingway)

Never mistake action for meaningful action.
                                (Hugo Kubinyi,2004)

Re: [Joelib-devel] smarts

[Joerg K. W. <we...@in...>]

And again,

see also:
joelib.desc.types.SSKey3DS in method getFingerprintPatterns()

see also:
joelib/data/plain/*.contributions for MR, PSA, LogP SMARTS contribution 
patterns.
They contain no $(...) environments, which is the most powerfull tool in 
SMARTS expression searching, so i recommend to use it heavily, although 
it is also time consuming.

Here are three more examples:
patty.addRule("[$([NH2]-c),ND1H3,ND2H2,ND3H1,ND2H1,$(Cl-[C,c]),$(Br-[C,c]),$(I-[C,c])]",
             "donor");
         patty.addRule("[$(N#C-[C,c]),OD1X1,OD2X2,ND3X3,ND2X2]", 
"acceptor");
         patty.addRule("[$([NH2]-C),$([OH]-C),$([OH]-c)]", "donoracceptor");

So, now you've nearly all public examples in JOELib collected in these 
mails.
I've also a SMARTS testing environment with side effect testing when 
changing the atom typer. Very usefull, for changing the expert system. 
I've used it also to resolve the aromaticity typer problems without 
loosing generalization ability.

So i will publish these classes and all molecular structures with the 
next release, so please be patient. Once a time if already submitted 
these classes to the JOELib/OpenBabel mailing list, but i will it make 
more official and write some documentation for this pretty piece of code.

Kind, regards, Joerg

> Hi,
> 
> sorry, replace all
> ,=or statement
> ;=and statement
> 
>> Hi naji,
>>
>> so here are my guesses, no gurantee that the SMART patterns work. So 
>> please resend the tested/working patterns.
>>
>>>> C= (participating in two fused benzene rings)
>>
>>
>> c~[#6]~c two benzene rings connected via aliphatic/aromatic via any bonds
>> or general aromatic rings ?
>> [r;a]~[#6]~[r;a]
>>
>>>> C= (participating in two fused non-benzene rings)
>>
>>
>> [r;!a]~[#6]~[r;!a] two non-aromatic rings connected via 
>> aliphatic/aromatic via any bonds
>>
>>>> C= (participating in two fused  rings: one benzene ring and one 
>>>> non-benzene ring)
>>
>>
>> [r;!a]~[#6]~[r;a]
>>
>>> -CH=( participating in a benzene ring)
>>>
>>>> C= (the formal double bond and a formal single bond participating in 
>>>> a benzene ring)
>>
>>
>> [#6,a,Q2,$(*-*),$(*=*)] so two heavy weight neighbors and neighbour 
>> environments with $(...), here general aromatic ring.
>> You can also work with hybridisation states using ^2, see 
>> joelib/data/plain/*.txt for examples
>>
>>>> C= (two single bonds participating in a non- benzene ring)
>>
>>
>> with or without implicite hydrogens, so X or D or Q
>> http://www-ra.informatik.uni-tuebingen.de/software/joelib/tutorial/functionalities/mol-smarts.html 
>>
>> and reference and link to daylight tutorial.
>> [#6,!a,r,Q2,$(*-*),$(*-*)] so two heavy weight neighbours, here 
>> general non-aromatic ring
>>
>>> -CH=( participating in a non- benzene ring)
>>
>>
>> [#6,!a,r,Q2,$(*-*),$(*=*)]
>>
>>>> C= (the formal double bond and a formal single bond participating in 
>>>> a non-benzene ring)
>>
>>
>> difference to -CH= ?
>>
>>> =N- (participating in a ring )
>>
>>
>> [#7,r,Q2,$(*-*),$(*-*)] so assuming any ring, or only aliphatic ones ?
>>
>>>> NH(participating in a ring )
>>
>>
>> [#7,r,H1,$(*-*),$(*-*)]
>>
>>> -N< (participating in a ring )
>>> -N< (participating in two fused rings )
>>
>>
>> analogue to C, more or less.
>>
>>> Any suggestions?
>>
>>
>> Yes :-)
>>
>>> Thanks
>>
>>
>> Was me a pleasure.
>>
>> Kind regards, Joerg
> 
> 
> 


-- 
Dipl. Chem. Joerg K. Wegner
Center of Bioinformatics Tuebingen (ZBIT)
Department of Computer Architecture
Univ. Tuebingen, Sand 1, D-72076 Tuebingen, Germany
Phone: (+49/0) 7071 29 78970
Fax: (+49/0) 7071 29 5091
E-Mail: mailto:we...@in...
WWW:    http://www-ra.informatik.uni-tuebingen.de
--
Never mistake motion for action.
                                     (E. Hemingway)

Never mistake action for meaningful action.
                                (Hugo Kubinyi,2004)