Do not renumber residue id in protein input file about oddt HOT 9 CLOSED

Thanks again ! Note that the exact module name is rdk ( oddt.toolkit = oddt.toolkits.rdk )

from oddt.

Hi Fabrice,

You're using Open Babel as a backend, right?

The idx attribute corresponds to the index used in OB, not to the residue number. However, the original residue numbering isn't lost and you can access it using res.OBResidue.GetNum() (where res is a residue object).

@mwojcikowski Is that intended? I think we should change it (RDKit uses residue number as idx)

from oddt.

Hi Marta,
I was using OB backend but in fact I don't know why ! When we use oddt.toolkit.readfile, OB is the default backend ?
Now I have switched the readfile to oddt.toolkits.rdk.readfile and RDKit seem to be the new backend with residue number as idx. This is exactetly what I want, perfect !
Thanks

from oddt.

If you have both RDKit and Open Babel installed, the latter is used by default.
You can change this behavior by setting ODDT_TOOLKIT env variable to rdkit, or by putting oddt.toolkit = oddt.toolkits.rdkit inside your script.

from oddt.

Sure, I think we should change the indexing in OB, although the idx property is taken from OpenBabel's GetIdx vs GetNum. Maybe we should make a distinction between them and create Residue.number property and fix Residue.idx as an 0-based index for RDKit.

from oddt.

Sounds reasonable

from oddt.

Hi,
In version 0.5 of oddt I was using
oddt.toolkit = oddt.toolkits.rdk #  force rdkit backend
By default oddt was not renumbering residue id in protein input file.
Since version oddt 0.6 with same rdkit backend the default behavior is now to renumbering.

How can I force oddt 0.6 (rdkit) to keep again the original numbering (including gap) ?

Could you provide some code example

Tanks

from oddt.

ODDT 0.6 introduced residue number property which maps to PDB numbering. idx which was previously that number is used as an consecutive index to match that of OpenBabel. So if I understand correctly all you need to do is to use Residue.number instead of Residue.idx (see the docs). If you have further problems please open new issue with a bit of code showing the changed behavior.

from oddt.

Here is a bit of code that prints PDB numbering with

import oddt
oddt.toolkit = oddt.toolkits.rdk #  force rdkit backend
mol = next(oddt.toolkit.readfile('pdb', 'tests/data/dude/xiap/receptor_rdkit.pdb'))
for res in mol.residues:
    print(res.number)

If you want to use atom_dict you need to mark a molecule as a protein:

import oddt
oddt.toolkit = oddt.toolkits.rdk #  force rdkit backend
mol = next(oddt.toolkit.readfile('pdb', 'tests/data/dude/xiap/receptor_rdkit.pdb'))
mol.protein = True
print(mol.atom_dict['resnum'])

from oddt.