[Cdk-devel] [ cdk-Patches-2781465 ] GenerateFragments.java - a possible patch

[SourceForge.net <no...@so...>]

Patches item #2781465, was opened at 2009-04-26 05:21
Message generated for change (Comment added) made by egonw
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=320024&aid=2781465&group_id=20024

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Boryeu Mao (boryeumao)
>Assigned to: Christian Hoppe (chhoppe)
Summary: GenerateFragments.java - a possible patch

Initial Comment:
The current algorithm for enumerating Bemis-Murcko fragments in GenerateFragment.java works well when the ring fragments of a molecule are linearly arranged, but misses some fragments when a molecule is 'branched'.  For example, the list of murcko fragments for the 'four-rings.mol' (in the attached gzipped-tar archive) does not include one of all four rings.  A possible alternative is to enumerate the fragments recursively ('GenerateFragments.java.patch' in the attached archive).  Instead of directly generating murcko fragments, the double loop f and g in the original code (beginning at lines 157 and 162 respectively) are used to generate relevant data on linkers that are composed of path atoms only.  The information on no-ring linkers are then passed to the newly added method recursiveFragmentAdd().

----------------------------------------------------------------------

>Comment By: Egon Willighagen (egonw)
Date: 2009-04-27 07:39

Message:
Christian, can you please have a look at the recent bug report (with unit
test) and this patch?

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=320024&aid=2781465&group_id=20024