Modifications can't just remove atoms

[pckroon]

Currently, modifications can only act on residues where there are unexpected atoms (such as PTMs). However, in some cases it is desireable to only remove atoms. E.g. uncharged amine groups. The canonical form would probably be the charged form (RNH3+), meaning an extraneous hydrogen gets reconstructed. And since there are no unexpected atoms, PTM processing is not applied, meaning it doesn't get removed again.

Applying the PTM processing to all residues where atoms were added is probably a Bad Idea, since most pdbs don't have any hydrogens, causing it to be run for the entire protein.

[pckroon]

If you annotate your protein with a modification from the CLI this now works. Otherwise it's pretty ambiguous whether the atom is really missing/unresolved, or whether it's a modified residue.
Requires quick testing to make sure.