@C/N-term mods do not create in silico modified peptides

[Shinya-Watanabe]

Hi,

I am trying to create a in silico spectral library with custom mods, but any of @Any/Protein_C/N-term mods do not create modified peptides, even the default Acetyl@Protein_N-term. I am using jupyter notebook with python 3.8 and pip installed all dependencies. Here's the brief process and output.

var_mods = ['Ethanolamine@D', 'Ethanolamine@E','Ethanolamine@Protein_C-term','PS@D','PS@E','PS@Protein_C-term']
user_mods = {"Ethanolamine@Protein_C-term":{"composition":"H(5)C(2)N(1)", "modloss_composition":""},
             "PS@D":{"composition":"H(5)C(3)N(1)O(2)", "modloss_composition":""},
             "PS@E":{"composition":"H(5)C(3)N(1)O(2)", "modloss_composition":""},
             "PS@Protein_C-term":{"composition":"H(5)C(3)N(1)O(2)", "modloss_composition":""}
             }

var_mods += ['Acetyl@Protein_N-term', 'Oxidation@M']
peptdeep.settings.add_user_defined_modifications(user_mods=user_mods)
peptdeep.settings.global_settings['common']['user_defined_modifications'] = user_mods
peptdeep.settings.global_settings['library']['var_mods'] = var_mods

Peptdeep.settings.add_user_defined_modifications() did not add user_mods to the global_settings, so I added them directly.

I created the spectral library using the following code.

fasta_lib = PredictSpecLibFasta(
    model_mgr,
    protease=digest,
    charged_frag_types=frag_types,
    var_mods=var_mods,
    fix_mods=['Carbamidomethyl@C'],
    max_missed_cleavages=max_miss_cleave,
    max_var_mod_num=max_var_mods,
    peptide_length_max=max_pep_len,
    peptide_length_min=min_pep_len,
    precursor_charge_min=precursor_charge_min,
    precursor_charge_max=precursor_charge_max,
    precursor_mz_min=min_pep_mz,
    precursor_mz_max=max_pep_mz,
    decoy=None
)

fasta_lib.get_peptides_from_fasta_list(fasta_list)
fasta_lib.append_decoy_sequence()
fasta_lib.add_modifications()

fasta_lib.precursor_df['nAA'] = fasta_lib.precursor_df.sequence.str.len()
fasta_lib.precursor_df.sort_values('nAA', inplace=True)
fasta_lib.precursor_df.reset_index(drop=True, inplace=True)
fasta_lib.add_charge()

I checked if @Protein_C-term is in 'mods' column using below, and there was nothing. I also checked 'Acetyl@Protein_N-term' but could not find.

prot_c_term_filter = fasta_lib.precursor_df['mods'].str.contains('@Protein_C-term')
fasta_lib.precursor_df[prot_c_term_filter]

Best,
Shinya