user.abbrevgroup file ignored by structure checker

User 873a9ae9d0

26-05-2016 17:01:28

Hi,


I want to check/fix structures in a sdfile using the alias checker/fixer.


Some of the abbreviations assigned to the atom alias(es) in my sdfile are not present in the default abbreviations file (that seems to be part of a Marvin jar file).


I know the way to go is create a (tab separated) textfile with name 'user.abbrevgroup' and it seems that I need to copy that file in the chemaxon folder of the user home directory. Since I run win(8.1) this would be something like e.g.


c:\users\theusername\chemaxon\user.abbrevgroup


Is that assumption correct? At least this is my understanding reading the docs.


If the assumption it does not work for me. I created a user.abbrev file (tab separated) and I assume I use the correct format and copied the file in the users home chemaxon folder.


When I run structure checker it seems that the content of this file is ignored, means I can specify what I want in my custom abbrev file, structure checker does not pick up the content of the file.


Could one please tell me what I am doing wrong. Hope someone can help.


Any feedback is highly appreciated.


Cheers


hans-juergen


 

ChemAxon e08c317633

27-05-2016 11:23:37

Yes, your information is correct this is how it should work.


Could you upload your user.abbrevgroup file (or just a part of it), an example input structure with alias atom, and your structure checker configuration file? We will check it and get back to you. How do you use structure checker (desktop app, command line, MarvinSketch, other)?

User 873a9ae9d0

29-05-2016 10:54:47

Hi,


 


thanks for responding.


Getting a confirmation that the user.abbrevgroup file is in the correct location I looked into other aspects of why checker did not fix my structures.


In that process I found a problem in the molfile of my structures containing alias atoms. The alias text was not assigned to a Carbon atom, instead a R atom was used to assign the alias text. This is (at least through the eyes of a molfile reader) totally OK. But checker (fixer) does fail to fix such structures.


Question: (and I assume the answer is no) Is there a 'trick' I can use to process those structures?


If the answer is No, I assume what I need to do is to find R atoms with assigned alias text and replace those with Carbon atoms.


Fortunately the way I ultimately use the checker is via the Java API, i.e. I have programmatic access to the atom(s) used to assigned an alias text and can programmatically change those.


Not sure if the usage of R (or I assume any atom from the extended ptable [like e.g. X]) atoms to assign alias text is something that no one else does. If not, it would be an useful enhancement to the chemaxon checker/fixer to be able to deal with those cases.


 


Best regards


hans-juergen


 


 

ChemAxon e08c317633

30-05-2016 10:28:04

Alias checker can detect aliases on R atoms, however the ConvertAliasToGroupFixer cannot convert them to proper groups. As far as I remember we did this on purpose.


You can implement your own fixer which can fix R atoms with alias labels. The details are described in the developers guide.


I hope this helps.