Is there any standardised way of indexing atoms

[syedzayyan]

Hello All!

I was following the RDKit Cookbook and tried to draw Caffeine using the given code:

This is the result from that. The code I use is this:

def mol_with_atom_index(mol):
    for atom in mol.GetAtoms():
        atom.SetAtomMapNum(atom.GetIdx())
    return mol

Then I was reading this paper:

Now they index their atoms differently here in this paper: https://www.cell.com/structure/pdfExtended/S0969-2126(17)30192-2

Now, my question is there any standard practise when indexing atoms and what could be the possible code for indexing atoms the 'right' way.

[bp-kelley]

To my knowledge, unless you are (1) explicitly talking about using IUPAC rules for indexing and naming or (2) using something like the the RCSB/PDB monomer library for things like creating structure submissions, there is no conventional way to index atoms.

REFS:
https://www.cas.org/sites/default/files/documents/indexguideapp.pdf
https://www.rcsb.org/ligand/LIB

[bp-kelley]

I will also point out that SetAtomMap(0) removes the atom map, you probably want to have one indexed atoms.

[syedzayyan]

I was somewhat asking IUPAC indexing. I am assuming the paper was using IUPAC as well since it matches the pubchem IUPAC name. Thank you very much for the resources.

[bp-kelley]

You can get the IUPAC locants using something like opsin if you have the IUPAC names.

https://opsin.ch.cam.ac.uk

But these don't seem to line up with the indexes used above.