Problem in Rgroup Decomposition with disconnected Scaffold

40 views
Skip to first unread message

mederich

unread,
Feb 3, 2012, 12:03:26 PM2/3/12
to indigo-...@googlegroups.com

Hi everybody,

 

I have found a problem with Scaffold Rgroup Decomposition when I specify a disconnected Scaffold:

 

For example, when I use this disconnected Scaffold ("CCCNC.OCCCC") on my set of molecule (InputMols.sdf) I obtain a good Rgroup Scaffold decomposition (Correct_decomposition.jpg).

However, when I use this other disconnected Scaffold ("CCCN.COCCCC") on the same set of input molecule I obtain a wrong Rgroup Scaffold decomposition (Wrong_decomposition.jpg).

 

For me Indigo has inverted the second part of the disconnected scaffold.

 

I hope my explanations will be help you to understand the problem.

 

With best regards.

 

Médérich.

 

 

 

Médérich Besnard

Ariana Pharma

28, rue du Dr Finlay - 75015 Paris - France

Tel: +33 (0)1 44 37 17 05 - Fax: +33 (0)1 44 37 17 01

www.arianapharma.com

 

InputMols.sdf
Correct_Decomposition.jpg
Wrong_Decomposition.jpg

Savelyev Alexander

unread,
Mar 2, 2012, 7:54:14 AM3/2/12
to indigo-...@googlegroups.com

Dear Mederich,

I have reproduced the issue. It is not a bug, because in the second case the 'CCCN' scaffold component has a different embedding for the target molecule 'mol1'. I have attached the molecule with highlighted scaffold. The problem is related to the iteration matchings issue. The iteration issue is discussed in 'RGroup Decomposition' topic. Also the 'RGroup Decomposition new API' topic may be interesting.
In this example the correct match embedding should be selected, because, in other case, the decomposed scaffold would have a different form. I think, with new interface, it can be done by iterating all the matches and selecting the match with minimum RSites number (may be more difficult algorithm).

With best regards,
Alexander


Savelyev Alexander

unread,
Apr 13, 2012, 6:46:39 AM4/13/12
to indigo-...@googlegroups.com
Hello all,

The new indigo 1.1beta11 version with the RGroup Decomposition match iterating API is available:
http://ggasoftware.com/download/indigo_next

You can iterate over all the matches in a molecule and append a correct decomposition to the full scaffold. The following python script(the similar on java and c#) describes the possibility to select the a match with minimum RSites number (mentioned in the previous letter):

------------------------------------------------------------------------------------------------------
# prepate query scaffold
scaffold = indigo.loadQueryMolecule("CCCNC.OCCCC")

# init decomposition
deco = indigo.createDecomposer(scaffold)

# load molecule
mol = indigo.loadMolecule("CCCCOCCCCCNCCC")

# create deco item
item = deco.processMolecule(mol)

min_r = 100
selected_match=None
# loop over all the matches
for q_match in item.iterateDecompositions():

   rg_mol = q_match.decomposedMoleculeWithRGroups()
   # search match with minimum RGroup count
   if rg_mol.countRSites() < min_r:
      min_r=rg_mol.countRSites()
      selected_match=q_match
   # add selected match to the full scaffold
    deco.addDecomposition(selected_match)

# full scaffold with only one RSite in this example
full_scaf = deco.decomposedMoleculeScaffold()

--------------------------------------------------------------------------------------------------------------
With best regards,
Alexander
Reply all
Reply to author
Forward
0 new messages