Antechamber’s new “acdoctor” sometimes misdiagnoses valences and rejects perfectly fine compounds. I am currently working around the issue by turning off the doctor with the “-dr n” flag, but thought I would provide a couple of examples of bad diagnoses so that hopefully acdoctor can be improved.
The first is pretty simple. It believes a nitro group implies 5 bonds to the nitrogen rather than 4, e.g. N2 in this compound:
The second is a misdiagnosis of the valences in/near a ring. In the following compound it believes there is a double bond between C33 and C34 (giving C33 a valence of 5) when the double bond is actually between C34 and C38.
—Eric
Eric Pettersen
UCSF Computer Graphics Lab
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Sep 27 2017 - 13:00:03 PDT
