Merge pull request #505 from lschaffer2/master

Option for only allowing common/known PTMs in identifications

ProteoformSuiteGUI/DisplayObjects/DisplayQuantitativeValues.cs

@@ -51,7 +51,8 @@ public string Theoretical
  {
  get
  {
- return proteoform.linked_proteoform_references != null ? proteoform.linked_proteoform_references.First().accession.Split('_')[0] : "";
+ return proteoform as TopDownProteoform != null ? proteoform.accession.Split('_')[0] :
+ proteoform.linked_proteoform_references != null ? proteoform.linked_proteoform_references.First().accession.Split('_')[0] : "";
  }
  }
 

ProteoformSuiteGUI/DisplayObjects/DisplayTopDownProteoform.cs

@@ -38,6 +38,11 @@ public string Name
  get { return t.name; }
  }
 
+ public string geneID
+ {
+ get { return t.geneID; }
+ }
+
  public string Sequence
  {
  get { return t.sequence; }

ProteoformSuiteGUI/ProteoformFamilies.cs

@@ -72,6 +72,9 @@ public void InitializeParameterSet()
 
  cb_count_adducts_as_id.Checked = Sweet.lollipop.count_adducts_as_identifications;
 
+ cb_only_assign_common_known_mods.Checked = Sweet.lollipop.only_assign_common_or_known_mods;
+
+
  initialize_every_time();
  }
 
@@ -380,5 +383,10 @@ private void cmbx_empty_TextChanged(object sender, EventArgs e)
  }
 
  #endregion Private Methods
+
+ private void cb_only_assign_common_known_mods_CheckedChanged(object sender, EventArgs e)
+ {
+ Sweet.lollipop.only_assign_common_or_known_mods = cb_only_assign_common_known_mods.Checked;
+ }
  }
 }

ProteoformSuiteInternal/Lollipop.cs

@@ -948,7 +948,7 @@ public void change_peak_acceptance(DeltaMassPeak peak, bool accepted, bool add_a
  public static bool gene_centric_families = false;
  public static string preferred_gene_label = "";
  public int deltaM_edge_display_rounding = 2;
-
+ public bool only_assign_common_or_known_mods = true;
  public static string[] node_positioning = new string[]
  {
  "Arbitrary Circle",

ProteoformSuiteInternal/Mods/custom_mods.txt

@@ -2,35 +2,35 @@ ID Sulfate Adduct
 TG Any
 PP Anywhere.
 MM 97.967377
-MT Unlocalized
+MT Common
 CF H2 O4 S1
 //
 ID Acetone Artifact (Unconfirmed)
 TG Any
 PP Anywhere.
 MM 98.0732
-MT Unlocalized
+MT Common
 CF C6 H10 O1
 //
 ID Hydrogen Dodecyl Sulfate
 TG Any
 PP Anywhere.
 MM 266.15516
-MT Unlocalized
+MT Common
 CF H26 C12 S1 O4
 //
 ID N-terminal Acetyl
 TG K
 PP N-terminal.
 MM 42.010565
-MT Unlocalized
+MT Common
 CF H2 C2 O1
 //
 ID GPI anchor
 TG Any
 PP Anywhere.
 MM 123.00853
-MT Unlocalized
+MT Common
 CF H6 C2 N1 O3 P1
 DR RESID; AA0161.
 DR RESID; AA0162.

ProteoformSuiteInternal/Mods/intact_mods.txt

@@ -3,59 +3,45 @@ ID Water Adduct
 TG Any
 PP Anywhere.
 MM 18.010565
-MT Unlocalized
+MT Common
 CF H2O
 //
 ID Water loss
 TG Any
 PP Anywhere.
 MM -18.010565
-MT Unlocalized
+MT Common
 CF H-2 O-1
 //
 ID Oxidation
 TG Any
 PP Anywhere.
 MM 15.994915
-MT Unlocalized
+MT Common
 CF O1
 //
 ID Acetyl
 TG K
 PP Anywhere.
 MM 42.010565
-MT Unlocalized
+MT Common
 CF H2 C2 O1
 //
-ID Carbamidomethyl of C
-TG C
-PP Anywhere.
-MM 57.021464
-MT Unlocalized
-CF H3 C2 N1 O1
-//
 ID Methyl
 TG E or V or H or R or K or Q or N or S or D
 PP Anywhere.
 MM 14.015650
-MT Unlocalized
+MT Common
 CF H2 C
 //
 ID Phospho
 TG S or T or Y
 PP Anywhere.
 MM 79.966331
 NL 0 or 97.976896
-MT Unlocalized
+MT Common
 CF H1 O3 P1
 //
-ID Phospho, Phospho, Phospho
-TG S or T or Y
-PP Anywhere.
-MM 239.898993
-MT Unlocalized
-CF H3 O9 P3
-//
 ID Missed Monoisotopic (-1)
 TG Any
 PP Anywhere.

ProteoformSuiteInternal/Mods/stored_mods.modnames

@@ -377,9 +377,9 @@ Methyl on Q Methyl 1 False
 Methyl on N Methyl 1 False
 Methyl on S Methyl 1 False
 Methyl on D Methy 1 False
-Phospho on S Phospho on S 1 False
-Phospho on T Phospho on T 1 False
-Phospho on Y Phospho on Y 1 False
+Phospho on S Phospho 1 False
+Phospho on T Phospho 1 False
+Phospho on Y Phospho 1 False
 Phospho, Phospho, Phospho on S Phospho, Phospho, Phospho 3 False
 Phospho, Phospho, Phospho on T Phospho, Phospho, Phospho 3 False
 Phospho, Phospho, Phospho on Y Phospho, Phospho, Phospho 3 False
@@ -407,4 +407,4 @@ Missing Serine (S) Missing Serine (S) 1 False
 Missing Alanine (A) Missing Alanine (A) 1 False
 Missing Glycine (G) Missing Glycine (G) 1 False
 Oxidation of M Ox 1 False
-Unmodified Unmodified 1 False
+Unmodified Unmodified 0 False

ProteoformSuiteInternal/Proteoform.cs

@@ -97,7 +97,7 @@ public List<ExperimentalProteoform> identify_connected_experimentals(List<PtmSet
  null); //Experimental without theoretical reference
 
  List<PtmSet> possible_additions = r.peak.possiblePeakAssignments.Where(p => Math.Abs(p.mass - deltaM) <= 1).ToList(); // EE relations have PtmSets around both positive and negative deltaM, so remove the ones around the opposite of the deltaM of interest
- PtmSet best_addition = generate_possible_added_ptmsets(possible_additions, deltaM, mass_tolerance, all_mods_with_mass, theoretical_base, 1)
+ PtmSet best_addition = generate_possible_added_ptmsets(possible_additions, deltaM, mass_tolerance, all_mods_with_mass, theoretical_base, 1, true)
  .OrderBy(x => (double)x.ptm_rank_sum + Math.Abs(x.mass - deltaM) * 10E-6) // major score: delta rank; tie breaker: deltaM, where it's always less than 1
  .FirstOrDefault();
 
@@ -151,7 +151,7 @@ public List<ExperimentalProteoform> identify_connected_experimentals(List<PtmSet
  }
 
  public List<PtmSet> generate_possible_added_ptmsets(List<PtmSet> possible_peak_assignments, double deltaM, double mass_tolerance, List<ModificationWithMass> all_mods_with_mass,
- TheoreticalProteoform theoretical_base, int additional_ptm_penalty)
+ TheoreticalProteoform theoretical_base, int additional_ptm_penalty, bool final_assignment)
  {
  List<ModificationWithMass> known_mods = theoretical_base.ExpandedProteinList.SelectMany(p => p.OneBasedPossibleLocalizedModifications.ToList()).SelectMany(kv => kv.Value).OfType<ModificationWithMass>().ToList();
  List<PtmSet> possible_ptmsets = new List<PtmSet>();
@@ -161,20 +161,13 @@ public List<PtmSet> generate_possible_added_ptmsets(List<PtmSet> possible_peak_a
  List<ModificationWithMass> mods_in_set = set.ptm_combination.Select(ptm => ptm.modification).ToList();
 
  int rank_sum = additional_ptm_penalty * (set.ptm_combination.Sum(m => Sweet.lollipop.theoretical_database.unlocalized_lookup.TryGetValue(m.modification, out UnlocalizedModification x) ? x.ptm_count : 1) - 1); // penalize additional PTMs
-
  foreach (ModificationWithMass m in mods_in_set)
  {
  int mod_rank = Sweet.lollipop.theoretical_database.unlocalized_lookup.TryGetValue(m, out UnlocalizedModification u) ? u.ptm_rank : Sweet.lollipop.modification_ranks.TryGetValue(m.monoisotopicMass, out int x) ? x : Sweet.lollipop.mod_rank_sum_threshold;
 
- if (m.monoisotopicMass == 0)
- {
- rank_sum += mod_rank;
- continue;
- }
-
  bool could_be_m_retention = m.modificationType == "AminoAcid" && m.motif.ToString() == "M" && theoretical_base.begin == 2 && this.begin == 2 && !ptm_set.ptm_combination.Any(p => p.modification.Equals(m));
- bool motif_matches_n_terminus = begin >= 1 && begin - 1 < theoretical_base.sequence.Length && m.motif.ToString() == theoretical_base.sequence[begin - 1].ToString();
- bool motif_matches_c_terminus = end >= 1 && end - 1 < theoretical_base.sequence.Length && m.motif.ToString() == theoretical_base.sequence[end - 1].ToString();
+ bool motif_matches_n_terminus = this.begin - theoretical_base.begin >= 0 && this.begin - theoretical_base.begin < theoretical_base.sequence.Length && m.motif.ToString() == theoretical_base.sequence[this.begin - theoretical_base.begin].ToString();
+ bool motif_matches_c_terminus = this.end - this.begin >= 0 && this.end - this.begin < theoretical_base.sequence.Length && m.motif.ToString() == theoretical_base.sequence[this.end - this.begin].ToString();
 
  bool cannot_be_degradation = !motif_matches_n_terminus && !motif_matches_c_terminus;
  if (m.modificationType == "Missing" && cannot_be_degradation
@@ -187,11 +180,28 @@ public List<PtmSet> generate_possible_added_ptmsets(List<PtmSet> possible_peak_a
 
  bool could_be_n_term_degradation = m.modificationType == "Missing" && motif_matches_n_terminus;
  bool could_be_c_term_degradation = m.modificationType == "Missing" && motif_matches_c_terminus;
- rank_sum -= Convert.ToInt32(Sweet.lollipop.theoretical_database.variableModifications.Contains(m)); // favor variable modifications over regular modifications of the same mass
+
+ //if selected, going to only allow mods in Mods folder (type "Common"), Missing, Missed Monoisotopic, known mods for that protein, or Unmodified
+ if (Sweet.lollipop.only_assign_common_or_known_mods && final_assignment)
+ {
+ if (!(m.monoisotopicMass == 0 || m.modificationType == "Common" || could_be_m_retention || could_be_n_term_degradation || could_be_c_term_degradation || m.modificationType == "Deconvolution Error" || known_mods.Concat(Sweet.lollipop.theoretical_database.variableModifications).Contains(m)))
+ {
+ rank_sum = Int32.MaxValue;
+ break;
+ }
+ }
 
  // In order of likelihood:
  // 1. First, we observe I/L/A cleavage to be the most common, other degradations and methionine cleavage are weighted mid-level
  // 2. Missed monoisotopic errors are considered, but weighted towards the bottom. This should allow missed monoisotopics with common modifications like oxidation, but not rare ones. (handled in unlocalized modification)
+ if (m.monoisotopicMass == 0)
+ {
+ rank_sum += mod_rank;
+ continue;
+ }
+
+ rank_sum -= Convert.ToInt32(Sweet.lollipop.theoretical_database.variableModifications.Contains(m)); // favor variable modifications over regular modifications of the same mass
+
  if (could_be_m_retention || could_be_n_term_degradation || could_be_c_term_degradation)
  {
  rank_sum += Sweet.lollipop.mod_rank_first_quartile / 2;
@@ -204,7 +214,9 @@ public List<PtmSet> generate_possible_added_ptmsets(List<PtmSet> possible_peak_a
  }
  else
  {
- rank_sum += known_mods.Concat(Sweet.lollipop.theoretical_database.variableModifications).Contains(m) ?
+ rank_sum += known_mods.Concat(Sweet.lollipop.theoretical_database.variableModifications).Contains(m) ||
+ known_mods.Select(mod => Sweet.lollipop.theoretical_database.unlocalized_lookup.TryGetValue(mod, out UnlocalizedModification um) ? um.id : mod.id).Contains(Sweet.lollipop.theoretical_database.unlocalized_lookup.TryGetValue(m, out UnlocalizedModification um2) ? um2.id : m.id) 
+ ?
  mod_rank :
  mod_rank + Sweet.lollipop.mod_rank_first_quartile / 2; // Penalize modifications that aren't known for this protein and push really rare ones out of the running if they're not in the protein entry
  }
@@ -268,8 +280,26 @@ private void assign_pf_identity(ExperimentalProteoform e, PtmSet set, Proteoform
  e.ptm_set.ptm_combination.Remove(ptm);
  }
  e.ptm_set = new PtmSet(e.ptm_set.ptm_combination);
+
+
+ e.uniprot_mods = "";
+ foreach (string mod in e.ptm_set.ptm_combination.Select(ptm => Sweet.lollipop.theoretical_database.unlocalized_lookup.TryGetValue(ptm.modification, out UnlocalizedModification x) ? x.id : ptm.modification.id).ToList().Distinct().OrderBy(m => m))
+ {
+ //positions with mod
+ List<int> theo_ptms = theoretical_base.ExpandedProteinList.First().OneBasedPossibleLocalizedModifications.Where(p => p.Key >= e.begin && p.Key <= e.end &&
+ p.Value.Where(m => m as ModificationWithMass != null).Select(m => Sweet.lollipop.theoretical_database.unlocalized_lookup.TryGetValue(m as ModificationWithMass, out UnlocalizedModification x) ? x.id : m.id).Contains(mod)).Select(m => m.Key).ToList();
+ if (theo_ptms.Count > 0)
+ {
+ e.uniprot_mods += mod + " @ " + String.Join(", ", theo_ptms) + "; ";
+ }
+ if (e.ptm_set.ptm_combination.Select(ptm => Sweet.lollipop.theoretical_database.unlocalized_lookup.TryGetValue(ptm.modification, out UnlocalizedModification x) ? x.id : ptm.modification.id).Count(m => m == mod)
+ > theo_ptms.Count)
+ {
+ e.novel_mods = true;
+ }
+ }
  }
- //if already been assigned -- check if gene name != this gene name ==> ambiguous and same length path
+ //if already been assigned -- check if gene name != this gene name ==> ambiguous
  else if (!e.topdown_id && (e.gene_name.get_prefered_name(Lollipop.preferred_gene_label) != this.gene_name.get_prefered_name(Lollipop.preferred_gene_label)))
  {
  e.ambiguous = true;
@@ -283,23 +313,6 @@ private void assign_pf_identity(ExperimentalProteoform e, PtmSet set, Proteoform
  {
  e.gene_name.gene_names.Concat(this.gene_name.gene_names);
  }
-
- e.uniprot_mods = "";
- foreach (string mod in e.ptm_set.ptm_combination.Select(ptm => Sweet.lollipop.theoretical_database.unlocalized_lookup.TryGetValue(ptm.modification, out UnlocalizedModification x) ? x.id : ptm.modification.id).ToList().Distinct().OrderBy(m => m))
- {
- //positions with mod
- List<int> theo_ptms = theoretical_base.ExpandedProteinList.First().OneBasedPossibleLocalizedModifications.Where(p => p.Key >= e.begin && p.Key <= e.end &&
- p.Value.Where(m => m as ModificationWithMass != null).Select(m => Sweet.lollipop.theoretical_database.unlocalized_lookup.TryGetValue(m as ModificationWithMass, out UnlocalizedModification x) ? x.id : m.id).Contains(mod)).Select(m => m.Key).ToList();
- if (theo_ptms.Count > 0)
- {
- e.uniprot_mods += mod + " @ " + String.Join(", ", theo_ptms) + "; ";
- }
- if (e.ptm_set.ptm_combination.Select(ptm => Sweet.lollipop.theoretical_database.unlocalized_lookup.TryGetValue(ptm.modification, out UnlocalizedModification x) ? x.id : ptm.modification.id).Count(m => m == mod)
- > theo_ptms.Count)
- {
- e.novel_mods = true;
- }
- }
  }
 
  #endregion Private Methods

ProteoformSuiteInternal/ProteoformRelation.cs

@@ -87,7 +87,7 @@ public ProteoformRelation(Proteoform pf1, Proteoform pf2, ProteoformComparison r
  TheoreticalProteoform t = pf2 as TheoreticalProteoform;
  double mass_tolerance = t.modified_mass / 1000000 * Sweet.lollipop.mass_tolerance;
  List<PtmSet> narrower_range_of_candidates = candidate_sets.Where(s => Math.Abs(s.mass - delta_mass) < Sweet.lollipop.peak_width_base_et).ToList();
- candidate_ptmset = pf1.generate_possible_added_ptmsets(narrower_range_of_candidates, delta_mass, mass_tolerance, Sweet.lollipop.theoretical_database.all_mods_with_mass, t, Sweet.lollipop.mod_rank_first_quartile)
+ candidate_ptmset = pf1.generate_possible_added_ptmsets(narrower_range_of_candidates, delta_mass, mass_tolerance, Sweet.lollipop.theoretical_database.all_mods_with_mass, t, Sweet.lollipop.mod_rank_first_quartile, false)
  .OrderBy(x => x.ptm_rank_sum + Math.Abs(Math.Abs(x.mass) - Math.Abs(delta_mass)) * 10E-6) // major score: delta rank; tie breaker: deltaM, where it's always less than 1
  .FirstOrDefault();
  }