Wrong mass for 2M-H+2NH4

[mapio]

I suspect there is a problem in the parsing of the ion part of the adduct leading to an error in the computation of the mass for 2M-H+2NH4.

Steps to reproduce, if you run

import pandas as pd 
from msac.calculate_adduct_mz import *

adducts = pd.read_csv('msac/example_data/adduct_list_full.csv')
bug = adducts[adducts.adduct == '2M-H+2NH4']
atom_dict, mass_dict, all_atoms = get_atom_masses(bug)
ions2mass = get_adduct_masses(None, mass_dict, all_atoms)

then ions2mass contains {'-H+2NH4': 29.01397304109} that is incorrect. Looking at intermediate results, mass_dict contains

{'e': 0.00054857990924, 'H': 1.00782503223, 'NH': 15.01089903666}

even if NH mass is correct, in the ion the formula is actually NH4. So the mass in ions2mass corresponds to

- mass_dict['H'] + 2 * mass_dict['NH']

that is a wrong computation. It should be

- Formula('H').monoisotopic_mass + 2 * Formula('NH4').monoisotopic_mass

which leads to 35.060923234469996 instead of 29.01397304109.

The same error happens also in just another case (from adduct_list_full.csv), namely M-H+2H2O where mass_dict again reports just O instead of O2.

I'm working on a complete overhaul of the parsing code that should fix this. As soon as I'll be ready I'll submit a pull request.