Hi Sushil,
On May 8, 2014, at 12:26 PM, Sushil Mishra wrote:
> MCS algorithm seems to me unable to handle chiral carbons and it can not
> differentiate chiral changes in ligands.
That's correct. The MCS algorithm in RDKit doesn't consider chirality. While in
principle I think it would be possible to extend the current algorithm to
support it, it would require some extensive changes.
> Moreover, it also fails to differentiate between position of atoms in
> symmetrical positions.For example I have 6 atoms ring (C1, C2, C3, C4, C5,
> O5) with one -CH3 at C1 and another structure with -CH3 at C5. MCS can not
> differentiate such structures.
The MCS search returns a SMARTS pattern, in this case something like:
[#6]1~[#6]~[#6]~[#6]~[#6]~([~6])~[#8]1
I don't think there's any way for a SMARTS, or a least a non-recursive SMARTS,
to handle those other than symmetrically.
It may be possible, through isotope labeling, for you to define your own atom
classes, so that the C1 atom in one carbohydrate can only ever match the C1
atom in another.
What would you like for it to return instead, in order to get the information
you need?
Cheers,
Andrew
[email protected]
------------------------------------------------------------------------------
Is your legacy SCM system holding you back? Join Perforce May 7 to find out:
• 3 signs your SCM is hindering your productivity
• Requirements for releasing software faster
• Expert tips and advice for migrating your SCM now
http://p.sf.net/sfu/perforce
_______________________________________________
Rdkit-discuss mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
