public void get_theoretical_proteoforms(string current_directory)
{
//Clear out data from potential previous runs
foreach (ProteoformCommunity community in Sweet.lollipop.decoy_proteoform_communities.Values)
{
community.theoretical_proteoforms = new TheoreticalProteoform[0];
}
theoretical_proteins.Clear();
//Read the UniProt-XML and ptmlist
List <Modification> all_known_modifications = get_mods(current_directory);
Parallel.ForEach(Sweet.lollipop.get_files(Sweet.lollipop.input_files, Purpose.ProteinDatabase).ToList(), database =>
{
if (database.extension == ".xml")
{
lock (theoretical_proteins)
theoretical_proteins.Add(database, ProteinDbLoader.LoadProteinXML(database.complete_path, true, DecoyType.None, all_known_modifications, database.ContaminantDB, Sweet.lollipop.mod_types_to_exclude, out Dictionary <string, Modification> um).ToArray());
lock (all_known_modifications) all_known_modifications.AddRange(ProteinDbLoader.GetPtmListFromProteinXml(database.complete_path).Where(m => !Sweet.lollipop.mod_types_to_exclude.Contains(m.ModificationType)));
}
else if (database.extension == ".fasta")
{
lock (theoretical_proteins)
theoretical_proteins.Add(database, ProteinDbLoader.LoadProteinFasta(database.complete_path, true, DecoyType.None, database.ContaminantDB, ProteinDbLoader.UniprotAccessionRegex, ProteinDbLoader.UniprotFullNameRegex, ProteinDbLoader.UniprotFullNameRegex, ProteinDbLoader.UniprotGeneNameRegex,
ProteinDbLoader.UniprotOrganismRegex, out var dbErrors).ToArray());
}
});
Sweet.lollipop.modification_ranks = rank_mods(theoretical_proteins, variableModifications, all_mods_with_mass);
unlocalized_lookup = make_unlocalized_lookup(all_mods_with_mass.Concat(new List <Modification> {
new Ptm().modification
}));
load_unlocalized_names(Path.Combine(Environment.CurrentDirectory, "Mods", "stored_mods.modnames"));
//this is for ptmsets --> used in RELATIONS
all_possible_ptmsets = PtmCombos.generate_all_ptmsets(2, all_mods_with_mass, Sweet.lollipop.modification_ranks, Sweet.lollipop.mod_rank_first_quartile / 2).ToList();
for (int i = 2; i <= Math.Max(ptmset_max_number_of_a_kind, Sweet.lollipop.max_ptms); i++) // the method above doesn't make 2 or more of a kind, so we make it here
{
all_possible_ptmsets.AddRange(all_mods_with_mass.Select(m => new PtmSet(Enumerable.Repeat(new Ptm(-1, m), i).ToList(), Sweet.lollipop.modification_ranks, Sweet.lollipop.mod_rank_first_quartile / 2)));
}
//Generate lookup table for ptm sets based on rounded mass of eligible PTMs -- used in forming ET relations
possible_ptmset_dictionary = make_ptmset_dictionary();
make_theoretical_proteoforms();
}
public void get_theoretical_proteoforms(string current_directory)
{
if (!ready_to_make_database(current_directory))
{
return;
}
//Clear out data from potential previous runs
foreach (ProteoformCommunity community in SaveState.lollipop.decoy_proteoform_communities.Values)
{
community.theoretical_proteoforms = new TheoreticalProteoform[0];
}
theoretical_proteins.Clear();
//Read the UniProt-XML and ptmlist
List <ModificationWithLocation> all_known_modifications = SaveState.lollipop.get_files(SaveState.lollipop.input_files, Purpose.PtmList).SelectMany(file => PtmListLoader.ReadModsFromFile(file.complete_path)).ToList();
uniprotModifications = make_modification_dictionary(all_known_modifications);
Dictionary <string, Modification> um;
Parallel.ForEach(SaveState.lollipop.get_files(SaveState.lollipop.input_files, Purpose.ProteinDatabase).ToList(), database =>
{
lock (theoretical_proteins) theoretical_proteins.Add(database, ProteinDbLoader.LoadProteinXML(database.complete_path, false, all_known_modifications, database.ContaminantDB, SaveState.lollipop.mod_types_to_exclude, out um).ToArray());
lock (all_known_modifications) all_known_modifications.AddRange(ProteinDbLoader.GetPtmListFromProteinXml(database.complete_path).OfType <ModificationWithLocation>().Where(m => !SaveState.lollipop.mod_types_to_exclude.Contains(m.modificationType)));
});
foreach (string filename in Directory.GetFiles(Path.Combine(current_directory, "Mods")))
{
var new_mods = !filename.EndsWith("variable.txt") || SaveState.lollipop.methionine_oxidation ?
PtmListLoader.ReadModsFromFile(filename) :
new List <ModificationWithLocation>(); // Empty variable modifications if not selected
if (filename.EndsWith("variable.txt"))
{
variableModifications = new_mods.OfType <ModificationWithMass>().ToList();
}
if (filename.EndsWith("intact_mods.txt"))
{
List <double> old_mods = all_known_modifications.OfType <ModificationWithMass>().Select(m => m.monoisotopicMass).ToList();
new_mods = new_mods.OfType <ModificationWithMass>().Where(m => !old_mods.Contains(m.monoisotopicMass)); // get rid of the unlocalized mods if they're already present
}
all_known_modifications.AddRange(new_mods);
}
all_known_modifications = new HashSet <ModificationWithLocation>(all_known_modifications).ToList();
uniprotModifications = make_modification_dictionary(all_known_modifications);
all_mods_with_mass = uniprotModifications.SelectMany(kv => kv.Value).OfType <ModificationWithMass>().Concat(variableModifications).ToList();
SaveState.lollipop.modification_ranks = rank_mods(theoretical_proteins, variableModifications, all_mods_with_mass);
unlocalized_lookup = make_unlocalized_lookup(all_mods_with_mass.Concat(new List <ModificationWithMass> {
new Ptm().modification
}));
load_unlocalized_names(Path.Combine(Environment.CurrentDirectory, "Mods", "stored_mods.modnames"));
//Generate all two-member sets and all three-member (or greater) sets of the same modification (three-member combinitorics gets out of hand for assignment)
all_possible_ptmsets = PtmCombos.generate_all_ptmsets(Math.Min(2, SaveState.lollipop.max_ptms), all_mods_with_mass, SaveState.lollipop.modification_ranks, SaveState.lollipop.mod_rank_first_quartile / 2).ToList();
for (int i = 2; i < SaveState.lollipop.max_ptms + 1; i++)
{
all_possible_ptmsets.AddRange(all_mods_with_mass.Select(m => new PtmSet(Enumerable.Repeat(new Ptm(-1, m), i).ToList(), SaveState.lollipop.modification_ranks, SaveState.lollipop.mod_rank_first_quartile / 2)));
}
//Generate lookup table for ptm sets based on rounded mass of eligible PTMs -- used in forming ET relations
possible_ptmset_dictionary = make_ptmset_dictionary();
expanded_proteins = expand_protein_entries(theoretical_proteins.Values.SelectMany(p => p).ToArray());
aaIsotopeMassList = new AminoAcidMasses(SaveState.lollipop.carbamidomethylation, SaveState.lollipop.natural_lysine_isotope_abundance, SaveState.lollipop.neucode_light_lysine, SaveState.lollipop.neucode_heavy_lysine).AA_Masses;
if (SaveState.lollipop.combine_identical_sequences)
{
expanded_proteins = group_proteins_by_sequence(expanded_proteins);
}
expanded_proteins = expanded_proteins.OrderBy(x => x.OneBasedPossibleLocalizedModifications.Count).ToArray(); // Take on harder problems first to use parallelization more effectively
process_entries(expanded_proteins, variableModifications);
process_decoys(expanded_proteins, variableModifications);
if (SaveState.lollipop.combine_theoretical_proteoforms_byMass)
{
SaveState.lollipop.target_proteoform_community.theoretical_proteoforms = group_proteoforms_by_mass(SaveState.lollipop.target_proteoform_community.theoretical_proteoforms);
foreach (ProteoformCommunity community in SaveState.lollipop.decoy_proteoform_communities.Values)
{
community.theoretical_proteoforms = group_proteoforms_by_mass(community.theoretical_proteoforms);
}
}
}
public void EnterTheoreticalProteformFamily(string seq, ProteinWithGoTerms prot, IDictionary <int, List <Modification> > modifications, string accession, List <TheoreticalProteoform> theoretical_proteoforms, int decoy_number, IEnumerable <Modification> variableModifications)
{
List <TheoreticalProteoform> new_theoreticals = new List <TheoreticalProteoform>();
if (seq.Length > 3000 || seq.Any(s => !aaIsotopeMassList.ContainsKey(s)))
{
return;
}
//Calculate the properties of this sequence
double unmodified_mass = TheoreticalProteoform.CalculateProteoformMass(seq, new List <Ptm>());
int lysine_count = seq.Split('K').Length - 1;
bool check_contaminants = theoretical_proteins.Any(item => item.Key.ContaminantDB);
//Figure out the possible ptm sets
Dictionary <int, List <Modification> > possibleLocalizedMods = modifications.ToDictionary(kv => kv.Key, kv => new List <Modification>(kv.Value));
foreach (Modification m in variableModifications)
{
for (int i = 1; i <= prot.BaseSequence.Length; i++)
{
if (prot.BaseSequence[i - 1].ToString() == m.Target.ToString())
{
if (!possibleLocalizedMods.TryGetValue(i, out List <Modification> a))
{
possibleLocalizedMods.Add(i, new List <Modification> {
m
});
}
else
{
a.Add(m);
}
}
}
}
int ptm_set_counter = 1;
//if top-down protein sequence, only add PTMs from that top-down proteoforms (will happen in add_topdown_theoreticals method)
if (!prot.topdown_protein)
{
List <PtmSet> unique_ptm_groups = PtmCombos.get_combinations(possibleLocalizedMods, Sweet.lollipop.max_ptms, Sweet.lollipop.modification_ranks, Sweet.lollipop.mod_rank_first_quartile / 2, limit_triples_and_greater);
//Enumerate the ptm combinations with _P# to distinguish from the counts in ProteinSequenceGroups (_#G) and TheoreticalPfGps (_#T)
foreach (PtmSet ptm_set in unique_ptm_groups)
{
TheoreticalProteoform t =
new TheoreticalProteoform(
accession + "_P" + ptm_set_counter.ToString(),
prot.FullDescription + "_P" + ptm_set_counter.ToString() + (decoy_number < 0 ? "" : "_DECOY_" + decoy_number.ToString()),
seq,
(prot as ProteinSequenceGroup != null ? (prot as ProteinSequenceGroup).proteinWithGoTermList.ToArray() : new ProteinWithGoTerms[] { prot }),
unmodified_mass,
lysine_count,
ptm_set,
decoy_number < 0,
check_contaminants,
theoretical_proteins);
t.topdown_theoretical = prot.topdown_protein;
new_theoreticals.Add(t);
ptm_set_counter++;
}
}
add_topdown_theoreticals(prot, seq, accession, unmodified_mass, decoy_number, lysine_count, new_theoreticals, ptm_set_counter, Sweet.lollipop.modification_ranks, Sweet.lollipop.mod_rank_first_quartile / 2);
lock (theoretical_proteoforms) theoretical_proteoforms.AddRange(new_theoreticals);
}
public void get_theoretical_proteoforms(string current_directory)
{
if (!ready_to_make_database(current_directory))
{
return;
}
//Clear out data from potential previous runs
foreach (ProteoformCommunity community in Sweet.lollipop.decoy_proteoform_communities.Values)
{
community.theoretical_proteoforms = new TheoreticalProteoform[0];
}
theoretical_proteins.Clear();
//Read the UniProt-XML and ptmlist
var psiModDeserialized = Loaders.LoadPsiMod(Path.Combine(current_directory, "Mods", "PSI-MOD.obo.xml"));
Dictionary <string, int> formalChargesDictionary = Loaders.GetFormalChargesDictionary(psiModDeserialized);
List <Modification> all_known_modifications = Sweet.lollipop.get_files(Sweet.lollipop.input_files, Purpose.PtmList)
.SelectMany(file => PtmListLoader.ReadModsFromFile(file.complete_path, formalChargesDictionary, out List <(Modification, string)> filteredModificationsWithWarnings))
.ToList();
uniprotModifications = make_modification_dictionary(all_known_modifications);
Parallel.ForEach(Sweet.lollipop.get_files(Sweet.lollipop.input_files, Purpose.ProteinDatabase).ToList(), database =>
{
if (database.extension == ".xml")
{
lock (theoretical_proteins)
theoretical_proteins.Add(database, ProteinDbLoader.LoadProteinXML(database.complete_path, true, DecoyType.None, all_known_modifications, database.ContaminantDB, Sweet.lollipop.mod_types_to_exclude, out Dictionary <string, Modification> um).ToArray());
lock (all_known_modifications) all_known_modifications.AddRange(ProteinDbLoader.GetPtmListFromProteinXml(database.complete_path).Where(m => !Sweet.lollipop.mod_types_to_exclude.Contains(m.ModificationType)));
}
else if (database.extension == ".fasta")
{
lock (theoretical_proteins)
theoretical_proteins.Add(database, ProteinDbLoader.LoadProteinFasta(database.complete_path, true, DecoyType.None, database.ContaminantDB, ProteinDbLoader.UniprotAccessionRegex, ProteinDbLoader.UniprotFullNameRegex, ProteinDbLoader.UniprotFullNameRegex, ProteinDbLoader.UniprotGeneNameRegex,
ProteinDbLoader.UniprotOrganismRegex, out var dbErrors).ToArray());
}
});
foreach (string filename in Directory.GetFiles(Path.Combine(current_directory, "Mods")))
{
List <Modification> new_mods = !filename.EndsWith("variable.txt") || Sweet.lollipop.methionine_oxidation ?
PtmListLoader.ReadModsFromFile(filename, formalChargesDictionary, out List <(Modification, string)> filteredModificationsWithWarnings).ToList() :
new List <Modification>(); // Empty variable modifications if not selected
if (filename.EndsWith("variable.txt"))
{
variableModifications = new_mods;
}
all_known_modifications.AddRange(new_mods);
}
all_known_modifications = new HashSet <Modification>(all_known_modifications).ToList();
uniprotModifications = make_modification_dictionary(all_known_modifications);
all_mods_with_mass = uniprotModifications.SelectMany(kv => kv.Value).Concat(variableModifications).ToList();
Sweet.lollipop.modification_ranks = rank_mods(theoretical_proteins, variableModifications, all_mods_with_mass);
unlocalized_lookup = make_unlocalized_lookup(all_mods_with_mass.Concat(new List <Modification> {
new Ptm().modification
}));
load_unlocalized_names(Path.Combine(Environment.CurrentDirectory, "Mods", "stored_mods.modnames"));
//this is for ptmsets --> used in RELATIONS
all_possible_ptmsets = PtmCombos.generate_all_ptmsets(2, all_mods_with_mass, Sweet.lollipop.modification_ranks, Sweet.lollipop.mod_rank_first_quartile / 2).ToList();
for (int i = 2; i <= Math.Max(ptmset_max_number_of_a_kind, Sweet.lollipop.max_ptms); i++) // the method above doesn't make 2 or more of a kind, so we make it here
{
all_possible_ptmsets.AddRange(all_mods_with_mass.Select(m => new PtmSet(Enumerable.Repeat(new Ptm(-1, m), i).ToList(), Sweet.lollipop.modification_ranks, Sweet.lollipop.mod_rank_first_quartile / 2)));
}
//Generate lookup table for ptm sets based on rounded mass of eligible PTMs -- used in forming ET relations
possible_ptmset_dictionary = make_ptmset_dictionary();
make_theoretical_proteoforms();
}