flavoinformatics / flavindb Goto Github PK

Bug:
As of writing the code filter_scripts doesn't handle any alternative nomenclature at all. This means that if an isoalloxazine is labelled as "C4a" (C4-alpha) instead of "C4x" that an error will be logged and the atom will be ignored by the script altogether.

Effect:
While this results in incorrect data (since correct data is missing), I suspect that there may be cases where a collision with alternate nomenclature occurs and the van Der Waals' forces are incorrectly looked up and the calculation may be altogether wrong; however, given a tolerance of 0.2 angstroms I don't think this case is causing any inaccuracies in the data).

Solution:
makeshift patch: extend filter_scripts/physical_constants.py to detect or use this alternative nomenclature so that we aren't currently missing any data

higher quality patch: scan the PDB file before doing any computation, look for alternate naming in the residues and attempt to map them to known ones names.

Read about biopandas

Read about biopandas

Low priority:

Turn filter scripts and it's associated functions into a callable python module.

There are a ton of dependencies in that file. Do we actually need all of them?