You feed a single compound identifier into Print_Compound_to_Disambiguate.py
eg:
./Print_Compound_to_Disambiguate.py cpd03657
The script collects all the data associated with the compound and consolidates it into a single JSON object, i.e. in this case:
(cpd02592_Sam_Seaver_Reactions_Object.json)[Disambiguation/Objects/cpd02592_Sam_Seaver_Reactions_Object.json]
(it uses your git name).
You edit this object, and you change any value from true to false if it doesn't belong to the originating compound (you might have to create a new compound, seen at the bottom of the object, though the code attempts to find one in the current biochemistry)
Then you apply the curation you did in the object by running:
./Integrate_Disambiguated_Compound.py Objects/cpd02592_Sam_Seaver_Object.json
This script separates out the data into the two compounds. But, if it finds that the disambiguation affects any reactions, then you get a new object to curate within which all the reactions associated with the compounds are listed, and you should edit to determine which reaction is for which compound:
./Integrate_Disambiguated_Reactions.py Objects/cpd02592_Sam_Seaver_Reactions_Object.json
In the Structures folder, you have to re-run code for updating structures and formulas in the database
./List_ModelSEED_Structures.py
./Update_Compound_Structures_Formulas_Charge.py
and in the Biochemistry folder, you have to re-run
the code for updating names, aliases, reaction mass balancing
etc. Best to use the bash script Refresh_Database.sh:
Refresh/Refresh_Database.sh
Finally, you should do these last few checks:
- Use git diff to check how the changes propagated
- Run
Reprint_Biochemistry.pyscript in the Biochemistry folder to check that it won't change again - Add and commit the objects generated by the curation process into the repository
- Commit your changes and make a new PR for us to review