private String GetPtmCsv (Peptide peptide) {
String result = "";
List<PTM> ptms = new List<PTM>();
foreach( List<PTM> variant in peptide.Variants )
foreach( PTM ptm in variant )
if( !ptms.Contains(ptm) ) {
ptms.Add(ptm);
result += peptide.ToString()+"/"+ptm.ToString()+" ";
}
return result;
}
private bool CheckPeptide( Peptide f ) {
if( f.Proteins.Count == 0 )
return false;
if( Spectra.Count != 0 && (f.Psm == null || f.Psm.Count == 0) )
return false;
if( FilterDecoys && f.Decoy )
return false;
if (f.Sequence.Length < LengthThreshold)
return false;
if( Type == SourceType.Plgs && (int)f.Confidence < (int)PlgsThreshold )
return false;
return true;
}
/// <summary>
/// Removes invalid PSMs and their associated peptides (if neccessary).
/// </summary>
private void FilterPsms() {
if( Spectra.Count == 0 )
return;
List<Spectrum> spectra = new List<Spectrum>();
// Remove previous relations
foreach( Peptide f in Peptides )
f.Psm = new List<PSM>();
foreach( Spectrum spectrum in Spectra ) {
if( spectrum.Psm == null )
continue;
Spectrum tmp = null;
foreach( PSM psm in spectrum.Psm ) {
if( !CheckPsm(psm) )
continue;
if( tmp == null ) {
tmp = new Spectrum();
tmp.File = spectrum.File;
tmp.ID = spectrum.ID;
tmp.SpectrumID = spectrum.SpectrumID;
tmp.Psm = new List<PSM>();
}
tmp.Psm.Add(psm);
Peptide f = psm.Peptide;
if( !f.Psm.Contains(psm) )
f.Psm.Add(psm);
if( f.Confidence < psm.Confidence )
f.Confidence = psm.Confidence;
}
if( tmp != null )
spectra.Add(tmp);
}
Spectra = spectra;
}
private void LoadData( string xmlpath ) {
SortedList<int,string> SortedAccession = new SortedList<int, string>();
XmlDocument doc = new XmlDocument();
doc.Load( xmlpath );
// Load proteins
XmlNodeList proteins = doc.GetElementsByTagName( "PROTEIN" );
foreach( XmlElement element in proteins ) {
int id = int.Parse(element.GetAttribute("ID"));
if( SortedAccession.ContainsKey(id) )
continue;
string acc = element.GetElementsByTagName("ACCESSION")[0].InnerText;
SortedAccession.Add( id, acc );
if( m_SortedProteins.ContainsKey(acc) )
continue;
string entry = element.GetElementsByTagName("ENTRY")[0].InnerText;
string desc = element.GetElementsByTagName("DESCRIPTION")[0].InnerText.Replace('+',' ');
string seq = element.GetElementsByTagName("SEQUENCE")[0].InnerText.ToUpper();
Protein p = new Protein(m_pid++, entry, acc, desc, seq);
Proteins.Add( p );
m_SortedProteins.Add( acc, p );
}
// Load peptides
SortedList<int,Peptide> SortedPeptides = new SortedList<int, Peptide>();
XmlNodeList peptides = doc.GetElementsByTagName( "PEPTIDE" );
foreach( XmlElement element in peptides ) {
int id = int.Parse(element.GetAttribute("ID"));
int pid = int.Parse(element.GetAttribute("PROT_ID"));
int mid = 0;
if( PlgsThreshold != Peptide.ConfidenceType.NoThreshold )
mid = int.Parse(element.GetAttribute("QUERY_MASS_ID"));
string seq = element.GetAttribute("SEQUENCE").ToUpper();
Peptide f = new Peptide(id, seq);
XmlNodeList mods = element.GetElementsByTagName( "MATCH_MODIFIER" );
f.Runs.Add( m_Run );
foreach( XmlElement mod in mods ) {
PTM ptm = new PTM();
string[] strs = mod.GetAttribute("NAME").Split(new char[]{'+'});
ptm.Name = strs[0];
string str = mod.GetAttribute("POS");
ptm.Pos = str.Length == 0 ? -1 : int.Parse(str);
if( strs.Length > 1 )
ptm.Residues = strs[1];
f.AddPTM( ptm );
}
Protein p = null;
try {
p = m_SortedProteins[SortedAccession[pid]];
} catch {
Notify( "Peptide '" + id + "' references unknown protein '" + pid + "'" );
}
if( p != null ) {
p.Peptides.Add( f );
f.Proteins.Add( p );
if( !p.Sequence.Contains(f.Sequence) )
throw new ApplicationException( "Inconsistent sequence data" );
}
Peptides.Add( f );
if( PlgsThreshold != Peptide.ConfidenceType.NoThreshold )
SortedPeptides.Add(mid,f);
}
if( PlgsThreshold == Peptide.ConfidenceType.NoThreshold )
return;
// Scores
XmlNodeList scores = doc.GetElementsByTagName( "MASS_MATCH" );
foreach( XmlElement element in scores ) {
int id = int.Parse(element.GetAttribute("ID"));
double score = double.Parse(element.GetAttribute("SCORE"), m_Format);
SortedPeptides[id].Score = score;
}
}
private void LoadData( string xmlpath )
{
SortedList<int,string> SortedAccession = new SortedList<int, string>();
XmlDocument doc = new XmlDocument();
doc.Load( xmlpath );
// Load proteins
XmlNodeList proteins = doc.GetElementsByTagName( "PROTEIN" );
foreach( XmlElement element in proteins ) {
int id = int.Parse(element.GetAttribute("ID"));
if( SortedAccession.ContainsKey(id) )
continue;
string acc = element.GetElementsByTagName("ACCESSION")[0].InnerText;
SortedAccession.Add( id, acc );
if( m_SortedProteins.ContainsKey(acc) )
continue;
string entry = element.GetElementsByTagName("ENTRY")[0].InnerText;
string desc = element.GetElementsByTagName("DESCRIPTION")[0].InnerText.Replace('+',' ');
string seq = element.GetElementsByTagName("SEQUENCE")[0].InnerText.ToUpper();
Protein p = new Protein(m_pid++, entry, acc, desc, seq);
Proteins.Add( p );
m_SortedProteins.Add( acc, p );
}
// Load peptides
SortedList<int,Peptide> SortedPeptides = new SortedList<int, Peptide>();
XmlNodeList peptides = doc.GetElementsByTagName( "PEPTIDE" );
foreach( XmlElement element in peptides ) {
int id = int.Parse(element.GetAttribute("ID"));
int pid = int.Parse(element.GetAttribute("PROT_ID"));
int mid = 0;
if( PlgsThreshold != Peptide.ConfidenceType.NoThreshold )
mid = int.Parse(element.GetAttribute("QUERY_MASS_ID"));
string seq = element.GetAttribute("SEQUENCE").ToUpper();
Peptide f = new Peptide(id, seq);
XmlNodeList mods = element.GetElementsByTagName( "MATCH_MODIFIER" );
f.Runs.Add( m_Run );
foreach( XmlElement mod in mods ) {
PTM ptm = new PTM();
string[] strs = mod.GetAttribute("NAME").Split(new char[]{'+'});
ptm.Name = strs[0];
string str = mod.GetAttribute("POS");
ptm.Pos = str.Length == 0 ? -1 : int.Parse(str);
if( strs.Length > 1 )
ptm.Residues = strs[1];
f.AddPTM( ptm );
}
Protein p = null;
try {
p = m_SortedProteins[SortedAccession[pid]];
} catch {
Notify( "Peptide '" + id + "' references unknown protein '" + pid + "'" );
}
if( p != null ) {
p.Peptides.Add( f );
f.Proteins.Add( p );
if( !p.Sequence.Contains(f.Sequence) )
throw new ApplicationException( "Inconsistent sequence data" );
}
Peptides.Add( f );
if( PlgsThreshold != Peptide.ConfidenceType.NoThreshold )
SortedPeptides.Add(mid,f);
}
if( PlgsThreshold == Peptide.ConfidenceType.NoThreshold )
return;
// Scores
XmlNodeList scores = doc.GetElementsByTagName( "MASS_MATCH" );
foreach( XmlElement element in scores ) {
int id = int.Parse(element.GetAttribute("ID"));
double score = double.Parse(element.GetAttribute("SCORE"), m_Format);
SortedPeptides[id].Score = score;
}
}
/// <summary>
/// Loads a mzIdentML file
/// </summary>
protected override void Load( string mzid )
{
m_mzid = new mzidFile1_0();
m_mzid.Load( mzid );
// Proteins
SortedList<string,string> SortedAccession = new SortedList<string, string>();
foreach( PSIPIanalysissearchDBSequenceType element in m_mzid.ListProteins ) {
if( SortedAccession.ContainsKey(element.id) )
continue;
string acc = element.accession;
SortedAccession.Add( element.id, acc );
if( m_SortedProteins.ContainsKey(acc) )
continue;
FuGECommonOntologycvParamType cv;
cv = FuGECommonOntologycvParamType.Find( "MS:1001352", element.cvParam );
string entry = cv == null ? "" : cv.value;
cv = FuGECommonOntologycvParamType.Find( "MS:1001088", element.cvParam );
string desc = cv == null ? "" : cv.value;
string seq = element.seq;//.ToUpper();
Protein p = new Protein(m_pid++, entry, acc, desc, seq);
p.DBRef = element.id;
Proteins.Add( p );
m_SortedProteins.Add( acc, p );
}
// Peptides
SortedList<string,Peptide> SortedPeptides = new SortedList<string, Peptide>();
int id = 1;
foreach( PSIPIpolypeptidePeptideType element in m_mzid.ListPeptides ) {
string seq = element.peptideSequence;//.ToUpper();
Peptide f = new Peptide(id++, seq);
f.Confidence = Peptide.ConfidenceType.PassThreshold; // It will be filtered later if neccessary
SortedPeptides.Add( element.id, f );
f.Runs.Add( m_Run );
if( element.Modification != null )
foreach( PSIPIpolypeptideModificationType mod in element.Modification ) {
PTM ptm = new PTM();
ptm.Pos = mod.locationSpecified ? mod.location : -1;
if( mod.residues != null )
foreach( string residue in mod.residues )
ptm.Residues += residue;
foreach( FuGECommonOntologycvParamType param in mod.cvParam )
if( param.cvRef.Equals("UNIMOD") )
ptm.Name = param.name;
f.AddPTM( ptm );
}
Peptides.Add( f );
}
// Relations
if( m_mzid.Data.DataCollection.AnalysisData.SpectrumIdentificationList.Length != 1 )
throw new ApplicationException( "Multiple spectrum identification lists not supported" );
foreach( PSIPIanalysissearchSpectrumIdentificationResultType idres in
m_mzid.Data.DataCollection.AnalysisData.SpectrumIdentificationList[0].SpectrumIdentificationResult )
foreach( PSIPIanalysissearchSpectrumIdentificationItemType item in idres.SpectrumIdentificationItem ) {
if( !item.passThreshold )
continue;
Peptide f = SortedPeptides[item.Peptide_ref];
if( item.PeptideEvidence == null )
continue;
f.Confidence = Peptide.ConfidenceType.PassThreshold;
foreach( PSIPIanalysisprocessPeptideEvidenceType relation in item.PeptideEvidence ) {
Protein p = m_SortedProteins[SortedAccession[relation.DBSequence_Ref]];
if( f.Proteins.Contains(p) )
continue;
f.Names.Add( relation.DBSequence_Ref, relation.id );
p.Peptides.Add( f );
f.Proteins.Add( p );
}
}
}
private String GetPtmCsv(Peptide peptide)
{
String result = "";
List<PTM> ptms = new List<PTM>();
foreach( List<PTM> variant in peptide.Variants )
foreach( PTM ptm in variant )
if( !ptms.Contains(ptm) ) {
ptms.Add(ptm);
result += peptide.ToString()+"/"+ptm.ToString()+" ";
}
return result;
}
private bool CheckPeptide( Peptide f )
{
if( f.Proteins.Count == 0 )
return false;
if( Spectra.Count != 0 && (f.Psm == null || f.Psm.Count == 0) )
return false;
if( FilterDecoys && f.Decoy )
return false;
if (f.Sequence.Length < LengthThreshold)
return false;
if( Type == SourceType.Plgs && (int)f.Confidence < (int)PlgsThreshold )
return false;
return true;
}
protected bool CheckPsm( bool passThreshold, int rank, Peptide.ConfidenceType confidence, double score, string type )
{
if( RequirePassTh && !passThreshold )
return false;
if( RankThreshold != 0 && (rank == 0 || rank > RankThreshold) )
return false;
if( Type >= SourceType.mzIdentML110 && Type <= SourceType.mzIdentML120 && (int)confidence < (int)SeqThreshold )
return false;
if( type == "Mascot expectation value" && score > MascotThreshold )
return false;
if( type == "X!Tandem expect" && score > XTandemThreshold )
return false;
return true;
}
/// <summary>
/// Loads a mzIdentML file
/// </summary>
override protected void Load( string mzid ) {
m_mzid = new mzidFile1_0();
m_mzid.Load( mzid );
// Proteins
SortedList<string,string> SortedAccession = new SortedList<string, string>();
foreach( PSIPIanalysissearchDBSequenceType element in m_mzid.ListProteins ) {
if( SortedAccession.ContainsKey(element.id) )
continue;
string acc = element.accession;
SortedAccession.Add( element.id, acc );
if( m_SortedProteins.ContainsKey(acc) )
continue;
FuGECommonOntologycvParamType cv;
cv = FuGECommonOntologycvParamType.Find( "MS:1001352", element.cvParam );
string entry = cv == null ? "" : cv.value;
cv = FuGECommonOntologycvParamType.Find( "MS:1001088", element.cvParam );
string desc = cv == null ? "" : cv.value;
string seq = element.seq;//.ToUpper();
Protein p = new Protein(m_pid++, entry, acc, desc, seq);
p.DBRef = element.id;
Proteins.Add( p );
m_SortedProteins.Add( acc, p );
}
// Peptides
SortedList<string,Peptide> SortedPeptides = new SortedList<string, Peptide>();
int id = 1;
foreach( PSIPIpolypeptidePeptideType element in m_mzid.ListPeptides ) {
string seq = element.peptideSequence;//.ToUpper();
Peptide f = new Peptide(id++, seq);
f.Confidence = Peptide.ConfidenceType.PassThreshold; // It will be filtered later if neccessary
SortedPeptides.Add( element.id, f );
f.Runs.Add( m_Run );
if( element.Modification != null )
foreach( PSIPIpolypeptideModificationType mod in element.Modification ) {
PTM ptm = new PTM();
ptm.Pos = mod.locationSpecified ? mod.location : -1;
if( mod.residues != null )
foreach( string residue in mod.residues )
ptm.Residues += residue;
foreach( FuGECommonOntologycvParamType param in mod.cvParam )
if( param.cvRef.Equals("UNIMOD") )
ptm.Name = param.name;
f.AddPTM( ptm );
}
Peptides.Add( f );
}
// Relations
if( m_mzid.Data.DataCollection.AnalysisData.SpectrumIdentificationList.Length != 1 )
throw new ApplicationException( "Multiple spectrum identification lists not supported" );
foreach( PSIPIanalysissearchSpectrumIdentificationResultType idres in
m_mzid.Data.DataCollection.AnalysisData.SpectrumIdentificationList[0].SpectrumIdentificationResult )
foreach( PSIPIanalysissearchSpectrumIdentificationItemType item in idres.SpectrumIdentificationItem ) {
if( !item.passThreshold )
continue;
Peptide f = SortedPeptides[item.Peptide_ref];
if( item.PeptideEvidence == null )
continue;
f.Confidence = Peptide.ConfidenceType.PassThreshold;
foreach( PSIPIanalysisprocessPeptideEvidenceType relation in item.PeptideEvidence ) {
Protein p = m_SortedProteins[SortedAccession[relation.DBSequence_Ref]];
if( f.Proteins.Contains(p) )
continue;
f.Names.Add( relation.DBSequence_Ref, relation.id );
p.Peptides.Add( f );
f.Proteins.Add( p );
}
}
}
/// <summary>
/// Checks for the presence of the given peptide in the peptide list of the protein
/// </summary>
public bool HasPeptide( Peptide fo ) {
return Peptides.Contains( fo );
}
private SortedList<string, Peptide> LoadPeptides()
{
SortedList<string,Peptide> SortedPeptides = new SortedList<string, Peptide>();
int id = 1;
foreach( PeptideType pep in m_mzid.ListPeptides ) {
Peptide p = new Peptide( id++, pep.PeptideSequence );
p.Confidence = Peptide.ConfidenceType.NoThreshold;
SortedPeptides.Add( pep.id, p );
p.Runs.Add( m_Run );
if( pep.Modification != null )
foreach( ModificationType mod in pep.Modification ) {
PTM ptm = new PTM();
ptm.Pos = mod.locationSpecified ? mod.location : -1;
if( mod.residues != null )
foreach( string residue in mod.residues )
ptm.Residues += residue;
foreach( CVParamType param in mod.cvParam )
if( param.cvRef.Equals("UNIMOD") )
ptm.Name = param.name;
p.AddPTM( ptm );
}
p.DBRef = pep.id;
Peptides.Add( p );
}
return SortedPeptides;
}
private void GetPsmScore(
SpectrumIdentificationItemType item,
out double score, out string type, out Peptide.ConfidenceType confidence )
{
score = -1.0;
type = "N/A";
confidence = item.passThreshold ? Peptide.ConfidenceType.PassThreshold : Peptide.ConfidenceType.NoThreshold;
if( item.Items == null )
return;
foreach( AbstractParamType param in item.Items ) {
if( !(param is CVParamType) )
continue;
CVParamType cv = param as CVParamType;
if( cv.accession == "MS:1001155" ) {
score = double.Parse(cv.value, m_Format);
type = "ProteomeDiscoverer/SEQUEST Confidence XCorr";
if( score >= m_GreenTh[item.chargeState-1] )
confidence = Peptide.ConfidenceType.Green;
else if( score >= m_YellowTh[item.chargeState-1] )
confidence = Peptide.ConfidenceType.Yellow;
else
confidence = Peptide.ConfidenceType.Red;
break;
} else if( cv.accession == "MS:1001172" ) {
score = double.Parse(cv.value, m_Format);
type = "Mascot expectation value";
break;
} else if( cv.accession == "MS:1001330" ) {
score = double.Parse( cv.value, m_Format );
type = "X!Tandem expect";
break;
}
}
}
/// <summary>
/// Removes peptides with low score, duplicated (same sequence) or not voted (multirun)
/// </summary>
private void FilterPeptides() {
List<Peptide> peptides = new List<Peptide>();
int id = 1;
// Remove previous relations
foreach( Protein p in Proteins )
p.Peptides.Clear();
// Filters duplicated (same sequence) peptides
SortedList<string,Peptide> SortedPeptides = new SortedList<string, Peptide>();
foreach( Peptide f in Peptides ) {
// Low score peptide
if( !CheckPeptide(f) ) {
if( f.Psm != null )
foreach( PSM psm in f.Psm )
psm.Spectrum.Psm.Remove(psm);
continue;
}
// Duplicated peptide, new protein?
if( SortedPeptides.ContainsKey(f.Sequence) ) {
Peptide fo = SortedPeptides[f.Sequence];
if( (int)f.Confidence > (int)fo.Confidence )
fo.Confidence = f.Confidence;
if( !fo.Runs.Contains(f.Runs[0]) )
fo.Runs.Add(f.Runs[0]);
fo.AddVariant( f.LastVariant );
bool dp = false; // duplicated protein?, needed for PLGS
foreach( Protein po in fo.Proteins )
if( po.ID == f.Proteins[0].ID ) {
dp = true;
break;
}
if( !dp )
fo.Proteins.Add( f.Proteins[0] );
if( fo.Psm == null )
fo.Psm = f.Psm;
else if( f.Psm != null )
fo.Psm.AddRange(f.Psm);
if( fo.Psm != null )
foreach( PSM psm in fo.Psm )
psm.Peptide = fo;
// New peptide
} else {
f.ID = id++;
SortedPeptides.Add( f.Sequence, f );
peptides.Add( f );
}
}
// Vote peptides
if( RunsThreshold > 1 ) {
Peptides = new List<Peptide>();
foreach( Peptide f in peptides )
if( f.Runs.Count >= RunsThreshold )
Peptides.Add(f);
} else
Peptides = peptides;
// Asigns new peptides to proteins
foreach( Peptide f in Peptides )
foreach( Protein p in f.Proteins )
p.Peptides.Add(f);
}
private void WriteProteinDetails( TextWriter w, Protein p ) {
Tag tr = new Tag( "tr", true );
Tag td = new Tag( "td", "colspan" );
Tag th = new Tag( "th", "rowspan" );
Tag a = new Tag( "a", "href" );
int i;
w.WriteLine( "<table>\n<caption><a name=\""+p.Accession+"\"/>Protein "+p.Accession+"</caption>" );
//w.WriteLine( "<col width=\"10%\"/><col width=\"5%\"/><col width=\"10%\"/><col width=\"75%\"/>" );
w.WriteLine( tr.Render(th.Render("Name")+td.Render("3",p.EntryEx)) );
w.WriteLine( tr.Render(th.Render("Description")+td.Render("3",p.Desc)) );
w.WriteLine( tr.Render(th.Render("Sequence")+td.Render("3","<pre>"+p.ParseSeq(10)+"</pre>")) );
w.WriteLine( tr.Render(th.Render("Evidence")+td.Render("3",p.Evidence.ToString())) );
w.Write( tr+th.Render("Peptide list")+"<td colspan=\"3\">" );
if( p.Peptides.Count > 0 ) {
for( i = 0; i < p.Peptides.Count-1; i++ )
w.Write( a.Render("#"+p.Accession+"__"+p.Peptides[i].ID,p.Peptides[i].ToString()) + ", " );
w.Write( a.Render("#"+p.Accession+"__"+p.Peptides[i].ID,p.Peptides[i].ToString()) );
}
w.WriteLine( "</td>"+tr );
if( p.Peptides.Count == 0 ) {
w.WriteLine( "</table><br/>" );
return;
}
int rows = 8;
if( Spectra.Count != 0 )
rows++;
w.Write( tr+th.Render((p.Peptides.Count*rows).ToString(),"Peptides") );
bool first = true;
foreach( Peptide f in p.Peptides ) {
if( first )
first = false;
else
w.Write( tr.ToString() );
w.Write( th.Render(rows.ToString(),f.ToString()) );
w.WriteLine( th.Render("<a name=\""+p.Accession+"__"+f.ID+"\"/>Confidence")+td.Render(f.Confidence.ToString())+tr );
tr.Hold = true;
w.WriteLine( tr.Render(th.Render("Decoy")+td.Render(f.Decoy.ToString())) );
w.Write( tr+th.Render("Runs")+td );
for( i = 0; i < f.Runs.Count-1; i++ )
w.Write( f.Runs[i].ToString() + ", " );
w.WriteLine( f.Runs[i].ToString()+td+tr );
w.WriteLine( tr.Render(th.Render("Relation")+td.Render(f.Relation.ToString())) );
w.Write( tr+th.Render("Proteins")+td );
for( i = 0; i < f.Proteins.Count-1; i++ )
w.Write( a.Render("#"+f.Proteins[i].Accession,f.Proteins[i].EntryEx) + ", " );
w.WriteLine( a.Render("#"+f.Proteins[i].Accession,f.Proteins[i].EntryEx)+td.ToString()+tr );
w.WriteLine( tr.Render(th.Render("Sequence")+td.Render("<pre>"+f.Sequence+"</pre>")) );
w.WriteLine( tr.Render(th.Render("Position")+td.Render(f.GetPositions(p))) );
w.Write( tr+th.Render("PTMs")+td );
if( f.Variants.Count == 1 )
w.Write( Peptide.Variant2Str(f.LastVariant) );
else {
i = 1;
foreach( List<PTM> v in f.Variants )
w.Write( "Variant #"+(i++)+": "+Peptide.Variant2Str(v)+"<br/>" );
}
w.WriteLine( td.ToString()+tr );
if( Spectra.Count != 0 ) {
w.Write( tr+th.Render("PSMs")+td );
if( f.Psm != null ) {
for( i = 0; i < f.Psm.Count-1; i++ )
w.Write( a.Render("#PSM"+f.Psm[i].ID,f.Psm[i].ID.ToString()) + ", " );
w.WriteLine( a.Render("#PSM"+f.Psm[i].ID,f.Psm[i].ID.ToString()) );
}
w.WriteLine( td.ToString()+tr );
}
tr.Hold = false;
}
w.WriteLine( "</table><br/>" );
}
/// <summary>
/// Checks for the presence of the given peptide in the peptide list of the protein
/// </summary>
public bool HasPeptide( Peptide fo )
{
return Peptides.Contains( fo );
}