private ProteinDetectionHypothesisType ClonePDH(ProteinDetectionHypothesisType pdh)
{
ProteinDetectionHypothesisType clon = new ProteinDetectionHypothesisType();
clon.id = pdh.id;
clon.name = pdh.name;
clon.dBSequence_ref = pdh.dBSequence_ref;
clon.passThreshold = pdh.passThreshold;
clon.PeptideHypothesis = pdh.PeptideHypothesis;
clon.Items = pdh.Items;
return(clon);
}
private void UpdatePdhCvs(ProteinDetectionHypothesisType pdh, Protein.EvidenceType evidence)
{
List <AbstractParamType> list = new List <AbstractParamType>();
foreach (AbstractParamType item in pdh.Items)
{
if (!item.name.Contains("PAnalyzer") && !item.name.Contains("leading"))
{
list.Add(item);
}
}
CVParamType ev = new CVParamType(), ld = new CVParamType();
ev.cvRef = "PSI-MS";
ld.cvRef = "PSI-MS";
switch (evidence)
{
case Protein.EvidenceType.Conclusive:
ev.accession = "MS:1002213";
ev.name = "PAnalyzer:conclusive protein";
ld.accession = "MS:1002401";
ld.name = "leading protein";
break;
case Protein.EvidenceType.Indistinguishable:
ev.accession = "MS:1002214";
ev.name = "PAnalyzer:indistinguishable protein";
ld.accession = "MS:1002401";
ld.name = "leading protein";
break;
case Protein.EvidenceType.Group:
ev.accession = "MS:1002216";
ev.name = "PAnalyzer:ambiguous group member";
ld.accession = "MS:1002401";
ld.name = "leading protein";
break;
case Protein.EvidenceType.NonConclusive:
ev.accession = "MS:1002215";
ev.name = "PAnalyzer:non-conclusive protein";
ld.accession = "MS:1002402";
ld.name = "non-leading protein";
break;
default: // filtered
return;
}
list.Add(ld);
list.Add(ev);
pdh.Items = list.ToArray();
}
/// <summary>
/// Create an object using the contents of the corresponding MzIdentML object
/// </summary>
/// <param name="pdh"></param>
/// <param name="idata"></param>
public ProteinDetectionHypothesisObj(ProteinDetectionHypothesisType pdh, IdentDataObj idata)
: base(pdh, idata)
{
Id = pdh.id;
Name = pdh.name;
DBSequenceRef = pdh.dBSequence_ref;
PassThreshold = pdh.passThreshold;
PeptideHypotheses = new IdentDataList <PeptideHypothesisObj>(1);
if (pdh.PeptideHypothesis?.Count > 0)
{
PeptideHypotheses.AddRange(pdh.PeptideHypothesis, ph => new PeptideHypothesisObj(ph, IdentData));
}
}
// TODO: Support empty PDH in input mzid
protected virtual List<ProteinAmbiguityGroupType> BuildProteinDetectionList() {
int gid = 1;
SortedList<string,ProteinDetectionHypothesisType> list = new SortedList<string, ProteinDetectionHypothesisType>();
List<ProteinAmbiguityGroupType> groups = new List<ProteinAmbiguityGroupType>();
foreach( ProteinAmbiguityGroupType grp in m_mzid.Data.DataCollection.AnalysisData.ProteinDetectionList.ProteinAmbiguityGroup )
foreach( ProteinDetectionHypothesisType pdh in grp.ProteinDetectionHypothesis )
list.Add( pdh.dBSequence_ref, pdh );
foreach( Protein p in Proteins ) {
ProteinAmbiguityGroupType g = new ProteinAmbiguityGroupType();
CVParamType ev = new CVParamType();
ev.accession = "MS:1001600";
ev.cvRef = "PSI-MS";
ev.name = "Protein Inference Confidence Category";
switch( p.Evidence ) {
case Protein.EvidenceType.Conclusive:
ev.value = "conclusive"; break;
case Protein.EvidenceType.Indistinguishable:
ev.value = "indistinguishable"; break;
case Protein.EvidenceType.Group:
ev.value = "ambiguous group"; break;
case Protein.EvidenceType.NonConclusive:
ev.value = "non conclusive"; break;
default:
continue;
}
g.id = "PAG_" + gid; gid++;
if( p.Subset.Count == 0 ) {
//g.ProteinDetectionHypothesis.Add(list[p.DBRef]);
g.ProteinDetectionHypothesis = new ProteinDetectionHypothesisType[]{list[p.DBRef]};
g.Items = new CVParamType[]{ev};
} else {
List<ProteinDetectionHypothesisType> listpdh = new List<ProteinDetectionHypothesisType>();
foreach( Protein p2 in p.Subset ) {
ProteinDetectionHypothesisType pdh = list[p2.DBRef];
pdh.Items = new CVParamType[]{ev};
listpdh.Add( pdh );
}
g.ProteinDetectionHypothesis = listpdh.ToArray();
}
groups.Add( g );
}
return groups;
}
/// <summary>
/// Create an object using the contents of the corresponding MzIdentML object
/// </summary>
/// <param name="pdh"></param>
/// <param name="idata"></param>
public ProteinDetectionHypothesisObj(ProteinDetectionHypothesisType pdh, IdentDataObj idata)
: base(pdh, idata)
{
this._id = pdh.id;
this._name = pdh.name;
this.DBSequenceRef = pdh.dBSequence_ref;
this._passThreshold = pdh.passThreshold;
this._peptideHypotheses = null;
if (pdh.PeptideHypothesis != null && pdh.PeptideHypothesis.Count > 0)
{
this.PeptideHypotheses = new IdentDataList <PeptideHypothesisObj>();
foreach (var ph in pdh.PeptideHypothesis)
{
this.PeptideHypotheses.Add(new PeptideHypothesisObj(ph, this.IdentData));
}
}
}
private List <ProteinAmbiguityGroupType> BuildPags()
{
int gid = 1;
SortedList <string, ProteinDetectionHypothesisType> listPdh = new SortedList <string, ProteinDetectionHypothesisType>();
SortedList <string, ProteinAmbiguityGroupType> listPag = new SortedList <string, ProteinAmbiguityGroupType>();
List <ProteinAmbiguityGroupType> groups = new List <ProteinAmbiguityGroupType>();
// Collect existing PDHs
foreach (ProteinAmbiguityGroupType grp in m_mzid.Data.DataCollection.AnalysisData.ProteinDetectionList.ProteinAmbiguityGroup)
{
foreach (ProteinDetectionHypothesisType pdh in grp.ProteinDetectionHypothesis)
{
if (!listPdh.ContainsKey(pdh.dBSequence_ref))
{
listPdh.Add(pdh.dBSequence_ref, pdh);
}
}
}
// Build PAGs for every group except non-conclusive proteins
foreach (Protein p in Proteins)
{
if (p.Evidence == Protein.EvidenceType.NonConclusive || p.Evidence == Protein.EvidenceType.Filtered)
{
continue;
}
ProteinAmbiguityGroupType g = new ProteinAmbiguityGroupType();
g.id = "PAG_" + gid++;
if (p.Subset.Count == 0)
{
g.ProteinDetectionHypothesis = new ProteinDetectionHypothesisType[] { listPdh[p.DBRef] };
UpdatePdhCvs(g.ProteinDetectionHypothesis[0], p.Evidence);
listPag.Add(g.ProteinDetectionHypothesis[0].dBSequence_ref, g);
}
else
{
List <ProteinDetectionHypothesisType> tmp = new List <ProteinDetectionHypothesisType>();
foreach (Protein p2 in p.Subset)
{
ProteinDetectionHypothesisType pdh = listPdh[p2.DBRef];
UpdatePdhCvs(pdh, p2.Evidence);
tmp.Add(pdh);
listPag.Add(pdh.dBSequence_ref, g);
}
g.ProteinDetectionHypothesis = tmp.ToArray();
}
groups.Add(g);
}
// Include non-conclusive proteins in existing PAGs
bool include;
foreach (Protein p in Proteins)
{
if (p.Evidence != Protein.EvidenceType.NonConclusive)
{
continue;
}
ProteinDetectionHypothesisType pdh = listPdh[p.DBRef];
UpdatePdhCvs(pdh, Protein.EvidenceType.NonConclusive);
foreach (Peptide f in p.Peptides)
{
foreach (Protein t in f.Proteins)
{
include = true;
if (!listPag.ContainsKey(t.DBRef))
{
continue;
}
ProteinAmbiguityGroupType g = listPag[t.DBRef];
foreach (ProteinDetectionHypothesisType h in g.ProteinDetectionHypothesis)
{
if (GetBaseId(h) == GetBaseId(pdh))
{
include = false;
break;
}
}
if (include)
{
ProteinDetectionHypothesisType clon = ClonePDH(pdh);
clon.id = GetBaseId(clon) + "@" + g.id;
List <ProteinDetectionHypothesisType> tmp =
new List <ProteinDetectionHypothesisType>(g.ProteinDetectionHypothesis);
tmp.Add(clon);
g.ProteinDetectionHypothesis = tmp.ToArray();
}
}
}
}
return(groups);
}
private String GetBaseId(ProteinDetectionHypothesisType pdh)
{
int i = pdh.id.IndexOf('@');
return(i < 0 ? pdh.id : pdh.id.Substring(0, i));
}