User 8fb8115204
10-12-2005 15:46:37
Please check the attached file.
It seems a bug of template base clean.
ChemAxon d76e6e95eb
10-12-2005 20:08:33
This is not a bug, but a missing feature. Your template contains 10 totally equivalent carbon atoms and they can be matched match in many ways to your target molecule. So, you should probably enhance your template by disclosing bridged carbons, or by defining that your ring is part of the SSSR, but there are no such query features available at the moment. We will discuss internally what we could do to avoid similar problems.
ChemAxon a3d59b832c
11-12-2005 21:40:59
Hi Blizzard,
We will add a new condition to template matching in template based clean. The template match will only be allowed if there are no "bridging chains" in the target between any two template atoms. (This will be decided using shortest path-s between atoms.) With this extra condition, your target will not be cleaned using the given template, so its original coordinates will be kept.
This new condition will be on by default, but it will be switchable.
Best regards,
Szabolcs
User 8fb8115204
13-12-2005 06:01:09
Hello Gyuri, Szabolcs
Thank you for your replies.
To add the query feature (r10) at the template maybe a workaround.
How do you think this workaround?
Thanks
ChemAxon a3d59b832c
14-12-2005 07:46:19
Yes, it will work in most cases. The exceptions are when smaller rings are fused to the 10-membered ring. (r10 means the size of the smallest ring the atom is part of). So the problematic cases:
- If you put r10 on each atoms, even one fusion will prevent the template from matching. (first attachment)
- If you put r10 only on one atom, you are more likely to match, but it also fails if all atoms of the large ring is part of a fusion. (second attachment)
So the best would be to put on only one r10 for now.
Best regards,
Szabolcs