Re: [Cdk-user] Question on CDK and a small documentation

2018-10-29 Thread Kohulan Rajan 

Dear John,
Thanks for the explanation and inputs. I require these modifications in order 
to depict a the molecule without much complexity. I will remember the warnings 
and do accordingly.
Thanks a lot once again for your help.
Have a nice evening.
Kind regards,
Kohulan.R
--
Kohulan Rajan
PhD Student
Faculty of Chemistry and Geosciences
Institute of Inorganic and Analytical Chemistry - Cheminformatics and 
Chemometrics
Friedrich-Schiller-University
Lessingstraße 8
07743 Jena , Germany
Phone : +49 3641 948783
“It is our choices that show what we truly are, far more than our abilities.” - 
Albus Dumbledore Monday, 29 October 2018, 06:17PM +01:00 from John Mayfield  
john.wilkinson...@gmail.com :

>I am still curious why, this kind of dubious manipulation is how errors start 
>propagating. Anyways I can only ask twice :-). These are not depiction options 
>but you can achieve it by modifying the molecule.
>
>1) I think what you're asking is to get the "parent" molecule. You should 
>define a list of the salts (counter ions) you want to remove and just remove 
>them from the IAtomContainer. A common hack is to remove everything but the 
>largest component (OEChem has the amusingly name:  
>OETheFunctionFormerlyKnownAsStripSalts ). In CDK you would use the  
>ConnectivityChecker  sort by size and take the largest one. Note you may end 
>up with a non-neutral form if care is not taken and if there are two 
>components of the same size you should decide what to do.
>
>2) Remove the stereochemistry,  
>container.setStereoElements(Collections.emptyList()) ;   prior to generating 
>coordinates. Or if you already have coordinates, iterate over the bonds and  
>setStereo(IBond.Stereo.NONE) . Depending on what you're asking you may want to 
>set the bond order to single ( setOrder(IBond.Order.SINGLE) ) too, note this 
>will then mess up valence and so you'd have radicals, so you'd have to sort 
>that out etc. Again removing information from the molecule is not a good.___
Cdk-user mailing list
Cdk-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/cdk-user

Re: [Cdk-user] Question on CDK and a small documentation

2018-10-29 Thread Kohulan Rajan 
Dear John,

Is there any possibilities to turn off the counter ion depiction in CDK?


Can we use a just straight lines for all types of bonds, rather than wedged 
bonds and dashed style bonds?


Kind Regards,
~Kohulan.R

___

Kohulan Rajan
PhD Student
Faculty of Chemistry and Geosciences
Institute of Inorganic and Analytical Chemistry - Cheminformatics and 
Chemometrics
Friedrich-Schiller-University
Lessingstraße 8, 07743 Jena , Germany
http://cheminf.uni-jena.de
Phone : +49 3641 948783

“It is our choices that show what we truly are, far more than our abilities.” - 
Albus Dumbledore

From: John Mayfield
Sent: Monday, October 29, 2018 3:59 PM
To: kohulan.ra...@uni-jena.de; cdkuser
Subject: Re: Question on CDK and a small documentation


Please use the cdk-user mailing (cc'd) for such questions in future. Other 
people than me can help, and if someone has the same question they'll get to 
see the answer too (it's also archived).

Is there any possibilities to turn off the counter ion depiction in CDK and can 
we use a just lines for all types of bonds, rather than wedged bonds and dashed 
style bonds?

Yes and yes. But I'd like a bit more details, was there something wrong that 
needs fixing?

Under Bond Count Descriptor the documentation says that you can use parameter 
“a” for aromatic bond counts, but after CDK 2.0 There is a new class called 
AromaticBondCountdescriptor made available. So now a developer cannot use that 
parameter. Also now a developer can use “q” for quadruple bonds, which should 
be added in the API documentation.

Please feel free to add a patch via GitHub that updates the JavaDoc.
 
John


On Mon, 29 Oct 2018 at 14:39, Kohulan Rajan  wrote:
Dear John,
 
Hope you are doing well. I am Kohulan Rajan , currently a Ph.D student working 
Under Prof.C.Steinbeck.
 
This is regarding a question regarding CDK and a small documentation in the new 
release.
 
Question.
    Is there any possibilities to turn off the counter ion 
depiction in CDK, and can we use a just lines for all types of bonds, rather 
than wedged bonds and dashed style bonds?
 
Update,
 
Under Bond Count Descriptor the documentation says that you can use parameter 
“a” for aromatic bond counts, but after CDK 2.0 There is a new class called 
AromaticBondCountdescriptor made available. So now a developer cannot use that 
parameter.
Also now a developer can use “q” for quadruple bonds, which should be added in 
the API documentation.
 

 
Awaiting for your reply.
 
Kind Regards,
~Kohulan.R

___

Kohulan Rajan
PhD Student
Faculty of Chemistry and Geosciences
Institute of Inorganic and Analytical Chemistry - Cheminformatics and 
Chemometrics
Friedrich-Schiller-University
Lessingstraße 8, 07743 Jena , Germany
http://cheminf.uni-jena.de
Phone : +49 3641 948783

“It is our choices that show what we truly are, far more than our abilities.” - 
Albus Dumbledore
 

___
Cdk-user mailing list
Cdk-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/cdk-user

Re: [Cdk-user] Question on CDK and a small documentation

2018-10-29 Thread John Mayfield 
I am still curious why, this kind of dubious manipulation is how errors
start propagating. Anyways I can only ask twice :-). These are not
depiction options but you can achieve it by modifying the molecule.

1) I think what you're asking is to get the "parent" molecule. You should
define a list of the salts (counter ions) you want to remove and just
remove them from the IAtomContainer. A common hack is to remove everything
but the largest component (OEChem has the amusingly name:
OETheFunctionFormerlyKnownAsStripSalts
).
In CDK you would use the ConnectivityChecker

sort
by size and take the largest one. Note you may end up with a non-neutral
form if care is not taken and if there are two components of the same size
you should decide what to do.

2) Remove the stereochemistry,
*container.setStereoElements(Collections.emptyList())*; prior to generating
coordinates. Or if you already have coordinates, iterate over the bonds and
*setStereo(IBond.Stereo.NONE)*. Depending on what you're asking you may
want to set the bond order to single (*setOrder(IBond.Order.SINGLE)*) too,
note this will then mess up valence and so you'd have radicals, so you'd
have to sort that out etc. Again removing information from the molecule is
not a good.
___
Cdk-user mailing list
Cdk-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/cdk-user

Re: [Cdk-user] Question on CDK and a small documentation

2018-10-29 Thread John Mayfield 
Please use the *cdk-user* mailing (cc'd) for such questions in future.
Other people than me can help, and if someone has the same question they'll
get to see the answer too (it's also archived).

Is there any possibilities to turn off the counter ion depiction in CDK and
> can we use a just lines for all types of bonds, rather than wedged bonds
> and dashed style bonds?


Yes and yes. But I'd like a bit more details, was there something wrong
that needs fixing?

Under Bond Count Descriptor the documentation says that you can use
> parameter “a” for aromatic bond counts, but after CDK 2.0 There is a new
> class called AromaticBondCountdescriptor made available. So now a developer
> cannot use that parameter. Also now a developer can use “q” for quadruple
> bonds, which should be added in the API documentation.


Please feel free to add a patch via GitHub that updates the JavaDoc.

John


On Mon, 29 Oct 2018 at 14:39, Kohulan Rajan 
wrote:

> Dear John,
>
>
>
> Hope you are doing well. I am Kohulan Rajan , currently a Ph.D student
> working Under Prof.C.Steinbeck.
>
>
>
> This is regarding a question regarding CDK and a small documentation in
> the new release.
>
>
>
> Question.
>
> Is there any possibilities to turn off the counter ion
> depiction in CDK, and can we use a just lines for all types of bonds,
> rather than wedged bonds and dashed style bonds?
>
>
>
> Update,
>
>
>
> Under Bond Count Descriptor the documentation says that you can use
> parameter “a” for aromatic bond counts, but after CDK 2.0 There is a new
> class called AromaticBondCountdescriptor made available. So now a developer
> cannot use that parameter.
>
> Also now a developer can use “q” for quadruple bonds, which should be
> added in the API documentation.
>
>
>
>
>
> Awaiting for your reply.
>
>
>
> Kind Regards,
> ~Kohulan.R
>
> ___
>
> Kohulan Rajan
> PhD Student
> Faculty of Chemistry and Geosciences
> Institute of Inorganic and Analytical Chemistry - Cheminformatics and
> Chemometrics
> Friedrich-Schiller-University
> Lessingstraße 8, 07743 Jena , Germany
>
> http://cheminf.uni-jena.de
> Phone : +49 3641 948783
>
> “It is our choices that show what we truly are, far more than our
> abilities.” - Albus Dumbledore
>
>
>
___
Cdk-user mailing list
Cdk-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/cdk-user