bingo -> indigo-nosql
python
c#

Right now only reading is available

Right now we have new group created in Maven central. The following changes are required

• Artefact "indigo" should be moved from com.epam --> com.epam.indigo group
• Artefact "bingo" should be renamed bingo-->indigo-nosql

Right now it is hard to understand what embedding is performed for the transformation (if any). Add mapping similar to substructure mapping

add new method to IndigoObject:
bool isPossibleFischerProjection(String parameters)

should return true if a structure contains at least one Fischer projection, falseotherwise

Currently Reaction Product Enumeration module breakes the invariant of the molecules provided as an output. Moreover: the dearomatization applied to such molecule makes it impossible to correct it in client modules.
This is why the dearomatization should become an optional procedure in RPE.

Current version doesn't support SDF files with length more 4 GB.

Summary:
when converting SMILES into internal representation, indigo should avoid placing stereobonds into ring bonds. They should be placed into non-ring bonds, where possible. This is especially important when structures like mono- and poly- saccharides where any non-conventional depiction raises questions and bug reports, e.g. the message below.

Obviously reading SMILES and saving again into MOLfile should behave the same way, i.e. it is an issue of SMILES conversion and not of depiction.

-----------original request-----------------------------------
I am a graduate student of Kansas University.
First, I really thanks to your great program, indigo-toolkit.
I would like to let you know the one of the problem of depiction.
I attached a example below.

I was going to depict the one of important monosaccharides, N-acetyl nuraminic acid, but the 2d depiction show me a little difference with the one in PDB site in terms of their stereochemistry. One of the edges in carbohydrate ring is expressed as hashed wedges.

code using by indigo-toolkit is like this.

‘''
indigo = Indigo()
renderer = IndigoRenderer(indigo)
outputname = 'test'
indigo.setOption("render-output-format", 'png')
indigo.setOption('render-background-color', 1.0, 1.0, 1.0)
test = indigo.loadMolecule("CC(=O)N[C@@h]1C@@HCC@@(O[C@H]1C@HC@HCO)C(O)=O")
renderer.renderToFile(test, '%s.png' % outputname)
‘’'

Could you please check how I can draw like picture above?
I look forward to your reply.

Thank you.

Sincerely,
Sang-Jun Park

python indigo-release-libs.py --preset=linux64
cmake -G "Unix Makefiles" -DSUBSYSTEM_NAME=x64 -DNO_STATIC=TRUE /home/dmitry/Indigo/build_scripts/indigo-all
-- System-specific folder name: Linux
-- Subsystem-specific folder name: x64
-- Set CMAKE_BUILD_TYPE to Release
-- CMAKE_BUILD_TYPE=Release
-- Common output dir: /home/dmitry/Indigo/build/indigo_Unix_Makefiles_ReleaseSUBSYSTEM_NAME_x64_NO_STATIC_TRUE/dist/Linux/x64
-- **** Indigo ****
-- Using local ZLib library
-- Indigo version is 1.2.2beta-r1
-- **** Indigo-inchi ****
-- **** Indigo-renderer ****
-- Using local LibPNG library
-- Using local Pixman library
-- Using local Cairo library
-- **** Bingo ****
-- Configuring done
CMake Warning (dev) in /home/dmitry/Indigo/api/CMakeLists.txt:
Policy CMP0022 is not set: INTERFACE_LINK_LIBRARIES defines the link
interface. Run "cmake --help-policy CMP0022" for policy details. Use the
cmake_policy command to set the policy and suppress this warning.

Target "indigo-shared" has an INTERFACE_LINK_LIBRARIES property which
differs from its LINK_INTERFACE_LIBRARIES properties.

INTERFACE_LINK_LIBRARIES:

-lpthread;-lrt;z;tinyxml;inchi

LINK_INTERFACE_LIBRARIES:

-lpthread;z;tinyxml;inchi

This warning is for project developers. Use -Wno-dev to suppress it.

-- Generating done
-- Build files have been written to: /home/dmitry/Indigo/build/indigo_Unix_Makefiles_ReleaseSUBSYSTEM_NAME_x64_NO_STATIC_TRUE
[ 4%] Built target z
[ 13%] Built target common
[ 13%] Built target tinyxml
[ 21%] Built target inchi
[ 27%] Built target graph
[ 41%] Built target molecule
[ 47%] Built target reaction
[ 51%] Built target layout
[ 51%] Built target dlopen-test
[ 56%] Built target indigoObj
[ 56%] Built target indigo-shared
[ 57%] Built target indigo-c-test-shared
[ 57%] Built target indigo-inchi-Obj
[ 57%] Built target indigo-inchi-shared
[ 57%] Built target indigo-inchi-c-test-shared
[ 60%] Built target png
[ 66%] Built target pixman
[ 66%] Building C object indigo-renderer/cairo/CMakeFiles/cairo.dir/src/cairo-ft-font.c.o
In file included from /home/dmitry/Indigo/third_party/cairo-src/cairo/src/cairo-ft-private.h:40:0,
from /home/dmitry/Indigo/third_party/cairo-src/cairo/src/cairo-ft-font.c:46:
/home/dmitry/Indigo/third_party/cairo-src/cairo/src/cairo-ft.h:50:35: fatal error: fontconfig/fontconfig.h: No such file or directory
#include <fontconfig/fontconfig.h>
^
compilation terminated.
make[2]: *** [indigo-renderer/cairo/CMakeFiles/cairo.dir/src/cairo-ft-font.c.o] Error 1
make[1]: *** [indigo-renderer/cairo/CMakeFiles/cairo.dir/all] Error 2
make: *** [all] Error 2
Traceback (most recent call last):
File "indigo-release-libs.py", line 158, in
build_libs(sys.argv[1:])
File "indigo-release-libs.py", line 127, in build_libs
subprocess.check_call("make package {0}".format('VERBOSE=1' if args.buildVerbose else ''), shell=True)
File "/usr/lib/python2.7/subprocess.py", line 540, in check_call
raise CalledProcessError(retcode, cmd)
subprocess.CalledProcessError: Command 'make package ' returned non-zero exit status 2

UpdateCTestConfiguration from :/home/dmitry/Indigo/build/indigo_Unix_Makefiles_ReleaseSUBSYSTEM_NAME_x64_NO_STATIC_TRUE/DartConfiguration.tcl
Parse Config file:/home/dmitry/Indigo/build/indigo_Unix_Makefiles_ReleaseSUBSYSTEM_NAME_x64_NO_STATIC_TRUE/DartConfiguration.tcl
UpdateCTestConfiguration from :/home/dmitry/Indigo/build/indigo_Unix_Makefiles_ReleaseSUBSYSTEM_NAME_x64_NO_STATIC_TRUE/DartConfiguration.tcl
Parse Config file:/home/dmitry/Indigo/build/indigo_Unix_Makefiles_ReleaseSUBSYSTEM_NAME_x64_NO_STATIC_TRUE/DartConfiguration.tcl
Test project /home/dmitry/Indigo/build/indigo_Unix_Makefiles_ReleaseSUBSYSTEM_NAME_x64_NO_STATIC_TRUE
Constructing a list of tests
Done constructing a list of tests
Checking test dependency graph...
Checking test dependency graph end
test 1
Start 1: dlopen-test

1: Test command: /home/dmitry/Indigo/build/indigo_Unix_Makefiles_ReleaseSUBSYSTEM_NAME_x64_NO_STATIC_TRUE/dist/Linux/x64/lib/dlopen-test "/home/dmitry/Indigo/api/libs/shared/Linux/x64/libindigo.so" "/home/dmitry/Indigo/api/libs/shared/Linux/x64/libindigo-inchi.so" "/home/dmitry/Indigo/api/libs/shared/Linux/x64/libindigo-renderer.so" "/home/dmitry/Indigo/api/libs/shared/Linux/x64/libbingo.so"
1: Test timeout computed to be: 20
1/4 Test #1: dlopen-test ......................***Exception: SegFault 0.14 sec
test 2
Start 2: indigo-c-test-shared

2: Test command: /home/dmitry/Indigo/build/indigo_Unix_Makefiles_ReleaseSUBSYSTEM_NAME_x64_NO_STATIC_TRUE/dist/Linux/x64/lib/indigo-c-test-shared
2: Test timeout computed to be: 20
2/4 Test #2: indigo-c-test-shared .............***Exception: SegFault 0.12 sec
test 3
Start 3: indigo-inchi-c-test-shared

3: Test command: /home/dmitry/Indigo/build/indigo_Unix_Makefiles_ReleaseSUBSYSTEM_NAME_x64_NO_STATIC_TRUE/dist/Linux/x64/lib/indigo-inchi-c-test-shared
3: Test timeout computed to be: 20
3/4 Test #3: indigo-inchi-c-test-shared .......***Exception: SegFault 0.13 sec
test 4
Start 4: indigo-renderer-c-test-shared

4: Test command: /home/dmitry/Indigo/build/indigo_Unix_Makefiles_ReleaseSUBSYSTEM_NAME_x64_NO_STATIC_TRUE/dist/Linux/x64/lib/indigo-renderer-c-test-shared
4: Test timeout computed to be: 20
4/4 Test #4: indigo-renderer-c-test-shared ....***Exception: SegFault 0.11 sec

0% tests passed, 4 tests failed out of 4

Total Test time (real) = 0.51 sec

The following tests FAILED:
1 - dlopen-test (SEGFAULT)
2 - indigo-c-test-shared (SEGFAULT)
3 - indigo-inchi-c-test-shared (SEGFAULT)
4 - indigo-renderer-c-test-shared (SEGFAULT)
Errors while running CTest
Traceback (most recent call last):
File "indigo-release-libs.py", line 158, in
build_libs(sys.argv[1:])
File "indigo-release-libs.py", line 140, in build_libs
subprocess.check_call("ctest -V --timeout 20 -C %s ." % (args.config), shell=True)
File "/usr/lib/python2.7/subprocess.py", line 540, in check_call
raise CalledProcessError(retcode, cmd)
subprocess.CalledProcessError: Command 'ctest -V --timeout 20 -C Release .' returned non-zero exit status 8
dmitry@dmitry-vb:~/Indigo/build_scripts$

Right now R sites are saved as "Any" atoms but should be saved as R
rsite_issue.mol.zip

Layout... layout.. layout..

attached-data.mol

See
https://groups.google.com/forum/#!topic/indigo-bugs/T3_KERMGYho
For molecular structures with more than 5 components the procedure shows poor performance. Initial solution might be to calculate canonical smiles for each component and sort alphabetical order the array of smiles.

Remove static initialization in pka model (molecule_ionize.cpp)

• create molecule_pka_model.cpp and move MoleculePkaModel class
• change indigo methods signatures

// Method for building PKA standard model
CEXPORT int indigoBuildPkaModel (int max_level, float threshold);

// Method for building PKA custom model from user SD file
CEXPORT int indigoBuildCustomPkaModel (int max_level, float threshold, const char *filename);

//use standard model
CEXPORT int indigoIonize (int item, float pH, float pH_toll);
CEXPORT int indigoIonizeCustom (int model, int item, float pH, float pH_toll);

CEXPORT float * indigoGetAcidPkaValue (int model, int item, int atom, int level, int min_level);
CEXPORT float * indigoGetBasicPkaValue (int model, int item, int atom, int level, int min_level);

Please see the attached file test.mol.zip
Code to reproduce

      Indigo indigo = new Indigo();

      indigo.setOption("treat-x-as-pseudoatom", "true");
      IndigoRenderer renderer = new IndigoRenderer(indigo);

      IndigoObject mol = indigo.loadMolecule(structure);
      indigo.setOption("render-coloring", "true");
      indigo.setOption("render-stereo-style", "none");
      indigo.setOption("render-image-size", 400, 400);
      indigo.setOption("render-output-format", "png");
      indigo.setOption("render-background-color", "1,1,1");

      byte[] test = renderer.renderToBuffer(mol);
      try (FileOutputStream writer = new FileOutputStream(new File("test.png"))) {
         writer.write(test);
      }

The presentation at http://www.slideshare.net/NextMoveSoftware/substructure-search-faceoff
on slide 20 for the same database and same search query shows different number of search hits for Indigo NoSQL (498,427) and Bingo(500,865). The search query is simple enough to expect that the number of hits should be the same. Database is available from here https://www.emolecules.com/info/plus/download-database
It is necessary to reproduce this finding and confirm that the number of hits is different. If this is the case, it is necessary to find a reason and if this is a bug, fix it.

Right now if you add properties (e.g. identifier) to IndigoObject monomers, and use the monomers in Reaction Product Enumeration, the result reaction removes the properties.
The expected behavior is to keep properties

Currently there are problems with aromatization/dearomatization of molecules with unusual number of hydrogens.

Lets take a molecule: N1C(S)=C2C([NH]=CN2)=NC=1N. Please see that the atom 5 (zero-based numbering) has one single, one double bond and one additional hydrogen. Thus the connectivity of nitrogen is 4.

The first question is if this molecule is valid (it was taken from the file with tautomers, but the question is open).
Next, there will be a problem if we perform sequential dearomatize(aromatize(molecule)): we will not get the initial molecule, but the one with several unresolved aromatic bonds. So, one of these procedures doesn't work well (if we assume that the molecule is valid).
If we compare the results of aromatization of this molecule given by Indigo and Marvin, we can see the difference.
Indigo: [n]1c(N)[n]c2[nH]c[nH]c2c1S (all cycles aromatized)
Marvin: Nc1nc(S)c2NC=[NH]c2n1 (only one cycle is aromatized)

Simple ring systems like C1=CC=CC=C1 should be layouted in "vertical" manner (see screen-shot

new option should skip indexing structures if atom count is bigger than input

Implement an optional (should be an option to manually set a version) procedure in build scripts to create a version from latest tag + relevant number of commits to the current state (revision)
r
e.g.
indigo1.2.2beta r3

molecule_layout_macrocycles.h(290): error C2864: 'indigo::TriangleLattice::_sink' : only static const integral data members can be initialized within a class

SHA ea9a78a
by Valeriy Samoylov, 10/17/2014 10:24 AM
parent 4cc429d

olecule_layout_macrocycles.h(315): error C2864: 'indigo::AnswerField::ACCEPTABLE_ERROR' : only static const integral data members can be initialized within a class

SHA 4cc429d
by Valeriy Samoylov, 09/11/2014 05:41 AM
parent af1a6e3

There is an exception while dearomatization:

array: invalid index 1 (size=1)

For structures like: N1N=N[c]2=[c]1[c]1=c[c]1=[c]2NN=N1

From Zoubeyr Boudene
https://groups.google.com/forum/#!topic/indigo-bugs/oX66ZxMBs1U

requested molecules : carboxylic acids containing C double bonds:

select *
from
molecules m
where
m.smiles @ ('CX3[OX2H]','')::bingo.smarts AND
m.smiles @ ('C=C','')::bingo.smarts

example of results that don't contain C double bonds:

"CCCCCCCCCC@HCC(O)=O"
"CCCCCCCCC@HCCC(O)=O"
"CC(CO)C(O)=O"
"OC(=O)Cc1ccccc1"
"Cc1ccccc1C(O)=O"
"Cc1cc(ccc1)C(O)=O"
"Cc1ccc(cc1)C(O)=O"

I don't have this issue if the column is not indexed.

CML saver for reactions adds incorrect additional molecules. Code to reproduce

   rxn = indigo.loadReactionFromFile("reaction_for_cml.rxn")
   print(rxn.countProducts())
   rxn.saveCml("test_rxn.cml")
   rxn2 = indigo.loadReactionFromFile("test_rxn.cml")
   print(rxn2.countProducts())

Output: 1 2

see attached RXN
reaction_for_cml.rxn.zip

New methods are required (java,python,c#)

• Array<IndigoObject index> bingoIndexStructures(Array molecules, String properties)
• Array<IndigoObject match> bingoMatchStructures(IndigoObject query, Array molecules, String properties)

Should process input structures by batches and return calculated values. index object contains

• fingerprints
• cmf
• coordinates
• exact hash
  ...

CMake pack seems to be configured incorrectly for Bingo PostgreSQL
Please see
https://groups.google.com/forum/#!topic/indigo-general/w_LRX109pfY
for some details

Here are the requirements:
• Input is RXN file (V3000 type), please refer to the attachments for the sample files.
• RXN file will not contain multiple reactions or multi-step reactions. Only one reaction is expected.
• Conversion should preserve all properties specified in RXN file, such as: reaction name, ISIS, etc.
• Conversion should support reaction atom-to-atom mapping.
• Outcome of the conversion is CDXML file which can be open by both ChemBioDraw Ultra (starting from version 13) and ChemDraw Direct (https://chemdrawdirect-cdn.perkinelmer.com/1.5.1/demo/index.html#)
• When displayed, converted reaction should be vertically and horizontally center aligned and not divided by columns.

Nice to have
• Support of different fonts and styles

Example of MOL file with Data SGroups and PARENT key:

SRU1
  ACCLDraw03131513562D

  0  0  0     0  0            999 V3000
M  V30 BEGIN CTAB
M  V30 COUNTS 14 15 3 0 0
M  V30 BEGIN ATOM
M  V30 1 C 12.2108 -4.4993 0 0 
M  V30 2 C 12.4501 -3.6062 0 0 
M  V30 3 C 11.5719 -3.3709 0 0 
M  V30 4 C 11.3326 -4.264 0 0 CFG=3 
M  V30 5 C 10.5453 -4.7185 0 0 
M  V30 6 C 9.758 -4.264 0 0 
M  V30 7 O 8.9708 -4.7185 0 0 
M  V30 8 C 8.1835 -4.264 0 0 
M  V30 9 C 5.8236 -5.6323 0 0 
M  V30 10 C 6.6134 -6.0822 0 0 
M  V30 11 C 7.3962 -5.6282 0 0 
M  V30 12 C 7.3962 -4.7185 0 0 
M  V30 13 C 6.6114 -4.2644 0 0 
M  V30 14 C 5.8236 -4.7189 0 0 
M  V30 END ATOM
M  V30 BEGIN BOND
M  V30 1 1 1 4 
M  V30 2 1 2 1 
M  V30 3 1 2 3 
M  V30 4 1 3 4 
M  V30 5 1 5 4 
M  V30 6 1 6 5 
M  V30 7 1 7 6 
M  V30 8 1 8 7 
M  V30 9 1 12 8 
M  V30 10 1 14 9 
M  V30 11 2 9 10 
M  V30 12 1 10 11 
M  V30 13 2 11 12 
M  V30 14 1 12 13 
M  V30 15 2 13 14 
M  V30 END BOND
M  V30 BEGIN SGROUP
M  V30 1 SRU 1 ATOMS=(3 7 5 6) XBONDS=(2 8 5) BRKXYZ=(9 8.57 -4.94 0 8.57 -
M  V30 -4.03 0 0 0 0) BRKXYZ=(9 10.94 -4.03 0 10.94 -4.94 0 0 0 0) -
M  V30 CONNECT=HT LABEL=n 
M  V30 2 DAT 2 PARENT=1 FIELDNAME=MOLWT_DISTRIBUTION -
M  V30 FIELDDISP="   -2.4375    1.6250    DR    ALL  1    D  5" -
M  V30 FIELDDATA=1.05 
M  V30 3 DAT 3 PARENT=1 FIELDNAME=MOLWT_NUMBER_AVERAGE -
M  V30 FIELDDISP="    0.3438    1.7188    DR    ALL  1    A  5" -
M  V30 FIELDDATA=5000 
M  V30 END SGROUP
M  V30 END CTAB
M  END

There is a termination error while saving molecules with pseudo atoms into CDXML format

Marvin Documents format looks like a wrapper around CML and is read by Indigo just fine. However the format has extensions for saving S-group information which is currently ignored. Here is an example of such data.

http://lifescience.opensource.epam.com/indigo/build.html#python

Requirements
Cmake (minimum required version is 2.8)
Make
Python (version 2.7)
MSVC(VS2013 or later) compiler for Windows and gcc for Linux 'c

Example
python indigo-release-libs.py --preset=win64-2012
change on:
python indigo-release-libs.py --preset=win64-2013

unsigned layoutRaing(int mol);

molecule_ionize.h(37): error C2864: 'indigo::IonizeOptions::level' : only static const integral data members can be initialized within a class
molecule_ionize.h(38): error C2864: 'indigo::IonizeOptions::min_level' : only static const integral data members can be initialized within a class

the faling code:

or lib in libs:
       name = "indigo-libs-%s-%s-shared%s" % (version, lib, suffix)
       if exists(name + ".zip"):

it happens because of the new convention to use '-r'

Right now for pseudo atoms there are no charges rendered. Use this code as an example

String structure = "\n"
              + "  Ketcher 03101613542D 1   1.00000     0.00000     0\n"
              + "\n"
              + "  1  0  0     0  0            999 V2000\n"
              + "    4.5500   -3.6750    0.0000 X   0  0  0  0  0  0  0  0  0  0  0  0\n"
              + "M  CHG  1   1  -1\n"
              + "M  END\n";

      Indigo indigo = new Indigo();

      indigo.setOption("treat-x-as-pseudoatom", "true");
      IndigoRenderer renderer = new IndigoRenderer(indigo);

      IndigoObject mol = indigo.loadMolecule(structure);
      indigo.setOption("render-coloring", "true");
      indigo.setOption("render-stereo-style", "none");
      indigo.setOption("render-image-size", 400, 400);
      indigo.setOption("render-output-format", "png");
      indigo.setOption("render-background-color", "1,1,1");

      byte[] halide1 = renderer.renderToBuffer(mol);
      try (FileOutputStream writer = new FileOutputStream(new File("halide1.png"))) {
         writer.write(halide1);
      }

https://groups.google.com/forum/#!topic/indigo-general/PJ4IA36a0LA

Add version parameter to indigo-depict tool
https://groups.google.com/forum/#!topic/indigo-general/F4xPnNp2MjA

Add methods for r-group composition (scaffold based enumeration). See the attached as an example. The output should contain enumerated products from scaffold and monomers
Create new method

iter indigoScaffoldProductEnumerate(mol, params)

• mol - a molecule with RGroups
• params - string
• iter - returns iterator for the enumerated products

I'm trying to use the Indigo API through ffi on OS X:

https://en.wikipedia.org/wiki/Foreign_function_interface

My target language is JavaScript (nodejs 5.6.0), and I'm using the node-ffi package.

I define my interface like this:

var libpath = './libindigo.dylib'
var lib = ffi.Library(indigoPath, {
  'indigoLoadMoleculeFromString': ['uint64', ['string']]
};

Loading a valid SMILES works fine:

var mol = lib.indigoLoadMoleculeFromString('c1ccccc1');

However, the node process exits when loading an invalid SMILES:

var mol = lib.indigoLoadMoleculeFromString('foobar');

libc++abi.dylib: terminating with uncaught exception of type indigo::SmilesLoader::Error
Abort trap: 6

So far I haven't found a way for node-ffi to catch errors thrown by a native library to allow them to be processed by the JavaScript interpreter.

Given that apparent limitation, is it possible to prevent Indigo from throwing errors in the first place? I though that's what the return value mechanism defined in indigo.h was for.

According to presentation https://github.com/rdkit/UGM_2015/blob/master/Presentations/RDKitHELM_RogerSayle.pdf (plots on pages 14-17) Indigo canonicalization algorithm does not scale well. Consider implementing algorithm from http://pubs.acs.org/doi/abs/10.1021/acs.jcim.5b00543 and https://github.com/rdkit/UGM_2015/blob/master/Presentations/Canonicalization_NadineSchneider.pdf