CM5 charges from XTB ASE interface

[leelasd]

Is your feature request related to a problem? Please describe.

Is there a way to get CM5 charges instead of Mulliken charges from xtb-ase interface? Any help would be highly appreciated.

Describe the solution you'd like

I have used the command line executable to generate CM5 charges, but it would be great to include that in python API.

Additional context

I have published on CM5 charges and their utility for modeling organic drug-like molecules. See the link below.
https://pubs.acs.org/doi/10.1021/acs.jctc.5b00414
Adding this feature would help seamlessly integrate XTB with force fields such as OpenFF

[awvwgk]

Implementing the CM5 correction is rather straight-forward (see grimme-lab/multicharge#7). Not sure what the best option is to make them available in the API, though.

[leelasd]

Thank you for the quick response @awvwgk ! Any suggestions on where to get started ?