Open Babel

On Apr 21, 2011, at 10:00 AM, remy d1 wrote:

> obgrep -c 'Cc1nnc(C)[nH]1' sdf-1-3500-match-triazole-result1.sdf
> 0

The main problem is the use of the explicit 'H' in the pattern. In your structure, the "lone" nitrogen (i.e., not the nn motif) has a carbon bond. A SMARTS pattern with an "H" requests that there be an explicit hydrogen on that atom. Of course there isn't one.

obgrep -c 'Cc1nnc(C)n1' triazole.sdf
1

> Moreover, on debian linux, I have seen that openbabel 2.2.3 includes libopenbabel.so.3 but obgrep needs libopenbabel.so.4

obgrep should be included in your Open Babel package. But in any case, libopenbabel.so.4 corresponds to v2.3.0, which you must have installed from another source. So you should have libopenbabel.so.4 if you have an obgrep which links to it.

(In any case, we strongly recommend the most recent release -- it has many fixes over 2.2.x, including both stereochemistry and aromaticity improvements.)

Hope that helps,
-Geoff

Hello

Thank you very much, Geoffrey.

Your message is very helpfull.

Regards,

Remy.

2011/4/21 Geoffrey Hutchison <geoff@...>

>
> On Apr 21, 2011, at 10:00 AM, remy d1 wrote:
>
> > obgrep -c 'Cc1nnc(C)[nH]1' sdf-1-3500-match-triazole-result1.sdf
> > 0
>
> The main problem is the use of the explicit 'H' in the pattern. In your
> structure, the "lone" nitrogen (i.e., not the nn motif) has a carbon bond. A
> SMARTS pattern with an "H" requests that there be an explicit hydrogen on
> that atom. Of course there isn't one.
>
> obgrep -c 'Cc1nnc(C)n1' triazole.sdf
> 1
>
> > Moreover, on debian linux, I have seen that openbabel 2.2.3 includes
> libopenbabel.so.3 but obgrep needs libopenbabel.so.4
>
> obgrep should be included in your Open Babel package. But in any case,
> libopenbabel.so.4 corresponds to v2.3.0, which you must have installed from
> another source. So you should have libopenbabel.so.4 if you have an obgrep
> which links to it.
>
> (In any case, we strongly recommend the most recent release -- it has many
> fixes over 2.2.x, including both stereochemistry and aromaticity
> improvements.)
>
> Hope that helps,
> -Geoff
>
>

After changing the bond order back to regular values (ie. bond order: 1, 2
3), when I try to detect the aromaticity, it still does not consider any of
the bonds to be aromatic.

mols[0]->BeginModify();
    FOR_BONDS_OF_MOL(b, mols[0]) {
        if (b->GetBondOrder() == 10) 
            b->SetBondOrder(1);
        else if( b->GetBondOrder() == 11)
            b->SetBondOrder(2);
        else if ( b->GetBondOrder() == 12 ) 
            b->SetBondOrder(3);
    }
    mols[0]->EndModify();


Is this not a correct way to do the begin and end modify function call?

Thanks,

Andi

Quinazoline is the molecule I am testing.
http://forums.openbabel.org/file/n3460579/Quinazoline.png 


--
View this message in context: http://forums.openbabel.org/Setting-Aromaticity-tp3416960p3460579.html
Sent from the General discussion mailing list archive at Nabble.com.

Downgraded from eigen3 to eigen2, and used:

cmake ../openbabel-2.3.0b2 -DPYTHON_BINDINGS=ON -DPERL_BINDINGS=ON
-DJAVA_BINDINGS=ON -DEIGEN2_INCLUDE_DIR=/usr/local/include/eigen2

make now proceeds fine, but make test shows a few worrisome errors:
	 44 - smartsparse_Test (Failed)
	 53 - pybindtest__pybel (Failed)
	  1 - align_Test (Timeout)
	  9 - sdfcansmi_Test (Failed)

>From perl, tried use Chemistry::OpenBabel;  but no go. 
Soo:
cd $OB_SOURCE/scripts/perl
perl Makefile.PL
make

Angry error stream:  using the cmake strategy of separate source directory
and build directory, babelconfig.h is in the build directory, not here in
source. Soo:
cp $OB_BUILD/include/openbabel/babelconfig.h  $OB_SRC/include/openbabel/
cd $OB_SOURCE/scripts/perl
perl Makefile.PL
make

Which compiles with just two warnings from ld. 
Running make test at this point suggests failure, but if you sudo make
install anyway, she works from within perl!!

Life is good. Hope these steps help someone else, even if it's just me when
installing on the next box :-)

--
View this message in context: http://forums.openbabel.org/Mac-OSX-10-6-6-compile-problems-tp3448049p3448442.html
Sent from the General discussion mailing list archive at Nabble.com.

On 14 April 2011 00:16, dave.nunn <david.s.nunn@...> wrote:
> Downgraded from eigen3 to eigen2, and used:
>
> cmake ../openbabel-2.3.0b2 -DPYTHON_BINDINGS=ON -DPERL_BINDINGS=ON
> -DJAVA_BINDINGS=ON -DEIGEN2_INCLUDE_DIR=/usr/local/include/eigen2
>
> make now proceeds fine, but make test shows a few worrisome errors:
>         44 - smartsparse_Test (Failed)
>         53 - pybindtest__pybel (Failed)
>          1 - align_Test (Timeout)
>          9 - sdfcansmi_Test (Failed)
>
> >From perl, tried use Chemistry::OpenBabel;  but no go.
> Soo:
> cd $OB_SOURCE/scripts/perl
> perl Makefile.PL
> make

No need to be running "perl Makefile.PL" yourself. It is run when you
do "make build". Do you run "make install"?

- Noel

> cmake ../openbabel-2.3.0b2 -DPYTHON_BINDINGS=ON -DPERL_BINDINGS=ON
> -DJAVA_BINDINGS=ON -DEIGEN2_INCLUDE_DIR=/usr/local/include/eigen2


Why are you using an old version of Open Babel? The latest release is 2.3.0 - final, not beta2. Not to mention that at some point fairly soon, there will be a 2.3.1 release.

-Geoff

Dear Paul,

> what I want to obtain: Add all possible hydrogens, i.e. every N.3 becomes
> N.4, and every COO- becomes COOH.

Ok - let's use a acid pH of 1, then (option -p1).

> For this purpose, I tried the following things:
> "babel -imol2 b_noH.mol2 -pX -omol2 b_noH_out.mol2"
> X was either 1, 7 or 14
>
> I tried the same for the MOL2 file in which the hydrogens were already in
> place for the non-polar atoms and the neutral nitrogen.
>
> Please find attached the corresponding MOL2 files.

Both files generate the same mol2 file for me:
babel -p1 b_noH.mol2 b_p1.mol2
(b_p1.mol2 attached)
N.4 with 3 H, O.co2 converted in =O (O.2) and OH (O.3).

If it's not the case for you, what's the output of
echo "NCC(=O)O" | babel -i smi -p1 -o smi
?

Check your OpenBabel version (should be 2.3.0) and be sure that there
is only one version installed.

By the way, I always use mol2 files with explicit hydrogens.

Regards,
Pascal

My concern is that different input protonation states of the same molecule do
not generate the same outputs for a given pH. E.g.:
echo "NCC(=O)O" | babel -i smi -p1 -o smi
[NH3+]CC(=O)O        #<< The expected result

echo "[NH3+]CC(=O)[O-]" | babel -i smi -p1 -o smi
[NH3+]CC(=O)[O-]    #<< Carboxylate didn't get reprotonated

echo "NCC(=O)O" | babel -i smi -p14 -o smi
NCC(=O)[O-]            #<< The expected result

echo "[NH3+]CC(=O)[O-]" | babel -i smi -p14 -o smi
[NH3+]CC(=O)[O-]    #<< Ammonium didn't get deprotonated

In these examples the parent (neutral) protonation state is handled
correctly, but anything with an initial charge is not de/re-protonated to
the desired pH. 

--
View this message in context: http://forums.openbabel.org/complete-protonation-tp3427279p3447662.html
Sent from the General discussion mailing list archive at Nabble.com.

On 12/04/2011 17:58, carlos ortega wrote:
> Greetings,
> I try to use babel to convert a multifile SDF to several files MOL2,
>
> I use this line:
>
> babel -isdf *.sdf -omol2 *.mol2 -m
>
> But the conversion put other name for the giles MOL2, and the file SDF has the
> name for each file MOL2.
>
> How I do to put the name?
> My file SDF has a lot of files, one million or more.

If you are using Windows
   babel  *.sdf  *.mol2
would convert each sdf file to a mol2 file with the same name.

With Linux, the shell replaces the second * character and you have to 
use instead
   babel  *.sdf  .mol2  -m

The newer, slightly different obabel interface is now recommended, so 
that the last command becomes
   obabel  *.sdf  -O .mol2  -m

Chris

On 12/04/2011 20:00, carlos ortega wrote:
> Greetings,
>
> I used the line
>
>     babel  *.sdf  .mol2  -m
>
> But, does not work.
>
> It put one file mol2 and the name is the same for the file SDF.

Sorry, I misunderstood what you were doing. Use
   babel  mymultifile.sdf  out.mol2  -m
which will produce a mol2 file for every molecule in the sdf file, 
starting with out1.mol2, out2.mol2 ...

Chris

>
> I use red hat 5.3 and i download the DAG packages for Red Hat Linux el5 x86_64,
> openbabel-2.1.1-1.el5.rf.x86_64.rpmHere
> http://rpmfind.net/linux/rpm2html/search.php?query=openbabel&submit=Search+...
>
> Can you help me?
>
> Thank you!
>
>
> Carlos Alberto Ortega Zuñiga
>
>
>
>
>
>
> ________________________________
> From: Chris Morley<c.morley@...>
> To: openbabel-discuss@...
> Sent: Tue, April 12, 2011 2:50:09 PM
> Subject: Re: [Open Babel] info about how convert from SDF multifile to mol2
>
> On 12/04/2011 17:58, carlos ortega wrote:
>> Greetings,
>> I try to use babel to convert a multifile SDF to several files MOL2,
>>
>> I use this line:
>>
>> babel -isdf *.sdf -omol2 *.mol2 -m
>>
>> But the conversion put other name for the giles MOL2, and the file SDF has the
>> name for each file MOL2.
>>
>> How I do to put the name?
>> My file SDF has a lot of files, one million or more.
>
> If you are using Windows
>     babel  *.sdf  *.mol2
> would convert each sdf file to a mol2 file with the same name.
>
> With Linux, the shell replaces the second * character and you have to
> use instead
>     babel  *.sdf  .mol2  -m
>
> The newer, slightly different obabel interface is now recommended, so
> that the last command becomes
>     obabel  *.sdf  -O .mol2  -m
>
> Chris
>
> ------------------------------------------------------------------------------
> Forrester Wave Report - Recovery time is now measured in hours and minutes
> not days. Key insights are discussed in the 2010 Forrester Wave Report as
> part of an in-depth evaluation of disaster recovery service providers.
> Forrester found the best-in-class provider in terms of services and vision.
> Read this report now!  http://p.sf.net/sfu/ibm-webcastpromo
> _______________________________________________
> OpenBabel-discuss mailing list
> OpenBabel-discuss@...
> https://lists.sourceforge.net/lists/listinfo/openbabel-discuss
>

On 12/04/2011 20:31, carlos ortega wrote:
> Thank you,
>
> this line work fine.
>
> But, is not possible put the name for each molecule that appear in the sdf file?
>
> I have about one million of molecules in the sdf file,
>
> each one have its name in the sdf file, and is some tedious search one name in
> one million of files, because the conversion puts other name.

I can see that this would be useful, but at present the molecule title 
cannot be used as the output file name with babel or obabel.

It would probably be possible using a simple Python script.

Chris
>
> ________________________________
> From: Chris Morley<c.morley@...>
> To: carlos ortega<caoz96@...>
> Sent: Tue, April 12, 2011 3:18:45 PM
> Subject: Re: [Open Babel] info about how convert from SDF multifile to mol2
>
> On 12/04/2011 20:00, carlos ortega wrote:
>> Greetings,
>>
>> I used the line
>>
>>      babel  *.sdf  .mol2  -m
>>
>> But, does not work.
>>
>> It put one file mol2 and the name is the same for the file SDF.
>
> Sorry, I misunderstood what you were doing. Use
>     babel  mymultifile.sdf  out.mol2  -m
> which will produce a mol2 file for every molecule in the sdf file,
> starting with out1.mol2, out2.mol2 ...
>
> Chris
>
>>
>> I use red hat 5.3 and i download the DAG packages for Red Hat Linux el5
> x86_64,
>> openbabel-2.1.1-1.el5.rf.x86_64.rpmHere
>> http://rpmfind.net/linux/rpm2html/search.php?query=openbabel&submit=Search+...
>>
>> Can you help me?
>>
>> Thank you!
>>
>>
>> Carlos Alberto Ortega Zuñiga
>>
>>
>>
>>
>>
>>
>> ________________________________
>> From: Chris Morley<c.morley@...>
>> To: openbabel-discuss@...
>> Sent: Tue, April 12, 2011 2:50:09 PM
>> Subject: Re: [Open Babel] info about how convert from SDF multifile to mol2
>>
>> On 12/04/2011 17:58, carlos ortega wrote:
>>> Greetings,
>>> I try to use babel to convert a multifile SDF to several files MOL2,
>>>
>>> I use this line:
>>>
>>> babel -isdf *.sdf -omol2 *.mol2 -m
>>>
>>> But the conversion put other name for the giles MOL2, and the file SDF has
> the
>>> name for each file MOL2.
>>>
>>> How I do to put the name?
>>> My file SDF has a lot of files, one million or more.
>>
>> If you are using Windows
>>      babel  *.sdf  *.mol2
>> would convert each sdf file to a mol2 file with the same name.
>>
>> With Linux, the shell replaces the second * character and you have to
>> use instead
>>      babel  *.sdf  .mol2  -m
>>
>> The newer, slightly different obabel interface is now recommended, so
>> that the last command becomes
>>      obabel  *.sdf  -O .mol2  -m
>>
>> Chris
>>
>> ------------------------------------------------------------------------------
>> Forrester Wave Report - Recovery time is now measured in hours and minutes
>> not days. Key insights are discussed in the 2010 Forrester Wave Report as
>> part of an in-depth evaluation of disaster recovery service providers.
>> Forrester found the best-in-class provider in terms of services and vision.
>> Read this report now!  http://p.sf.net/sfu/ibm-webcastpromo
>> _______________________________________________
>> OpenBabel-discuss mailing list
>> OpenBabel-discuss@...
>> https://lists.sourceforge.net/lists/listinfo/openbabel-discuss
>>

Hi,

>> But, is not possible put the name for each molecule that appear in the sdf file?

> I can see that this would be useful, but at present the molecule title
> cannot be used as the output file name with babel or obabel.
>
> It would probably be possible using a simple Python script.
>

Sure, probably the best way, IMHO.

Or split the sdf (output files should be the molecule name) (with e.g.
SplitSDFiles.pl from Mayachemtools; I have my own very basic perl
script too if you want to give it a try), and convert to mol2.
Or convert the multi sdf to multi mol2, and split the mol2 (I have a
basic perl script for that too - it should take less that one day for
1 million molecule, if I remember well, on my own PC).
Or split the sdf with obabel, rename the file (I have a tcl script if
the name is in a sdf field, or perl script which need an input file
with mol names).

But all that stuff was before I knew pybel, which can do it in one shoot:

#!/usr/bin/python
import pybel
for mol in pybel.readfile("sdf", "tmp.sdf"):
	filename = mol.OBMol.GetTitle()
	filename = filename + ".mol2"
	output = open(filename,"w") # erase existing file, if any
	output.write(mol.write("mol2"))
        output.close()

That's it.
Comments welcome, I'm always eager to learn how do it better, as
scripting is not my biggest skill :)

Regards,
Pascal

Le 2011-04-12 à 09:27, ndk@... a écrit :
> Hello!
> 
> I am very grateful to you for your help! I couldn’t understand what the
> problem was for a long time. I looked at your files and now it seems
> everything is OK.

  OK, great - and thanks for your files, this was a good way to test thoroughly 
the cif2cartesian conversion.
  Now we should probably indicate somewhere in the OB doc the official reference 
for the convention used. Where ?

	Vincent
-- 
Vincent Favre-Nicolin

CEA / INAC                              http://inac.cea.fr
Université Joseph Fourier        http://www.ujf-grenoble.fr

http://vincefn.net
ObjCryst & Fox : http://objcryst.sourceforge.net

On 12 April 2011 13:49, Vincent Favre-Nicolin
<vincefn@...> wrote:
> Le 2011-04-12 à 09:27, ndk@... a écrit :
>> Hello!
>>
>> I am very grateful to you for your help! I couldn’t understand what the
>> problem was for a long time. I looked at your files and now it seems
>> everything is OK.
>
>  OK, great - and thanks for your files, this was a good way to test thoroughly
> the cif2cartesian conversion.
>  Now we should probably indicate somewhere in the OB doc the official reference
> for the convention used. Where ?

I would put it in the cifformat.cpp in the Description. This text is
used is display by "babel -H", and is also used to create the online
docs (and book!).

- Noel

Le 12/04/2011 17:03, Noel O'Boyle a écrit :
> On 12 April 2011 13:49, Vincent Favre-Nicolin
>>  Now we should probably indicate somewhere in the OB doc the official reference
>> for the convention used. Where ?
> 
> I would put it in the cifformat.cpp in the Description. This text is
> used is display by "babel -H", and is also used to create the online
> docs (and book!).

  OK, done - I hope it's ok because I can't manage to link openbabel
HEAD now - some issue with undefined references to dlopen and dlerror in
src/dlhandler_unix.cpp (?).

	Vincent

Dear Daniel，
             Thank you very much for your help.  Anyway, people can do such
work by themselves according to the IUPAC rules.
             I hope Openbabel can develop such a module, but I know it is
very difficult.
shen



2011/4/12 dan2097 <dl387@...>

>
> qiancheng shen wrote:
> >
> > Thank you very much!! But all these tools are not suitable for
> me......Any
> > other ideas?
> >
> I don't believe there any Open Source structure to name tools. Whilst the
> vocabulary for name to structure and structure to name are the same the
> actual techniques involved are quite different so producing a structure to
> name tool would be a significant effort.
> Are you in academia? If so you can get free academic licenses for either
> OpenEye's Lexichem or ChemAxon's Structure to Name.
>
> If you have ChemScript with your ChemDraw license you can in principle use
> that to batch process moderate numbers of structures to names.
>
>
> Geoffrey Hutchison-2 wrote:
> >
> > Name to structure has a clear benefit for OPSIN and is technically an
> > easier problem: parse the chemical name into determining structure. There
> > may be many names for one structure, but that's not a big deal. OPSIN was
> > also used to solve an obvious problem -- machine parsing journal articles
> > for data mining.
> >
> Which problem is harder depends on what the goal of the project is IMO.
> Producing a systematic name for a compound whilst far from easy is not
> extremely difficult. Relatively few nomenclature procedures can describe a
> vast amount of chemistry unambiguously, if inelegantly. Producing "good"
> systematic names , however, is far harder considering all the different
> types of nomenclature that should ideally be applied to different types of
> chemistry and in different cases.
>
> Name to structure for strictly IUPAC names is moderately easy but for the
> general case of chemical names is somewhat of an unbouded problem due to
> human variations in nomenclature. Even within IUPAC nomenclature there are
> a
> lot of different allowed variations especially if you consider older
> nomenclature recommendations.
>
>
> Igor Filippov [Contr]-2 wrote:
> >
> > 6. CIR....
> > what's wrong with CIR?
> >
> The CIR can act as a name lookup service but it cannot name a compound
> which
> it has never encountered before.
>
> Daniel
>
> --
> View this message in context:
> http://forums.openbabel.org/Convert-Molecule-Structure-to-IUPAC-tp3424850p3443290.html
> Sent from the General discussion mailing list archive at Nabble.com.
>
>
> ------------------------------------------------------------------------------
> Forrester Wave Report - Recovery time is now measured in hours and minutes
> not days. Key insights are discussed in the 2010 Forrester Wave Report as
> part of an in-depth evaluation of disaster recovery service providers.
> Forrester found the best-in-class provider in terms of services and vision.
> Read this report now!  http://p.sf.net/sfu/ibm-webcastpromo
> _______________________________________________
> OpenBabel-discuss mailing list
> OpenBabel-discuss@...
> https://lists.sourceforge.net/lists/listinfo/openbabel-discuss
>