public void CalculateUniqueQValue(Dictionary <OptimalResultCondition, List <IIdentifiedSpectrum> > peptideBin)
{
foreach (var spectra in peptideBin.Values)
{
IdentifiedSpectrumUtils.CalculateUniqueQValue(spectra, scoreFunc, fdrCalc);
}
}
public void TestSameEngineDifferentParameters()
{
ClassificationOptions co = new ClassificationOptions();
co.ClassifyByCharge = true;
co.ClassifyByMissCleavage = true;
co.ClassifyByModification = true;
co.ModifiedAminoacids = "STY";
co.ClassifyByNumProteaseTermini = true;
var s1 = new MascotPeptideTextFormat().ReadFromFile(TestContext.CurrentContext.TestDirectory + "/../../../data/deisotopic.peptides");
IdentifiedSpectrumUtils.RemoveSpectrumWithAmbigiousAssignment(s1);
s1.ForEach(m => m.Tag = "deisotopic");
var s2 = new MascotPeptideTextFormat().ReadFromFile(TestContext.CurrentContext.TestDirectory + "/../../../data/deisotopic-top10.peptides");
IdentifiedSpectrumUtils.RemoveSpectrumWithAmbigiousAssignment(s2);
s2.ForEach(m => m.Tag = "deisotopic-top");
var all = s1.Union(s2).ToList();
var p1 = new List <IIdentifiedSpectrum>(all);
IdentifiedSpectrumUtils.KeepTopPeptideFromSameEngineDifferentParameters(p1, new ScoreFunction());
p1.ForEach(m => m.ClassificationTag = "deisotopic/deisotopic-top");
var bin1 = co.BuildSpectrumBin(p1);
var p2 = new List <IIdentifiedSpectrum>(all);
IdentifiedSpectrumUtils.KeepUnconflictPeptidesFromSameEngineDifferentParameters(p2, new ScoreFunction());
p2.ForEach(m => m.ClassificationTag = "deisotopic/deisotopic-top");
var bin2 = co.BuildSpectrumBin(p2);
bin2.ForEach(m =>
{
IdentifiedSpectrumUtils.KeepTopPeptideFromSameEngineDifferentParameters(m.Spectra, new ScoreFunction());
var n = bin1.Find(a => a.Condition.ToString().Equals(m.Condition.ToString()));
Assert.AreEqual(m.Spectra.Count, n.Spectra.Count);
//{
// if (m.Condition.ToString().Equals("deisotopic/deisotopic-top; Charge=2; MissCleavage=0; Modification=1; NumProteaseTermini=2"))
// {
// Assert.IsTrue(n.Spectra.Any(k => k.Query.FileScan.ShortFileName.Equals("20111128_CLi_v_4-2k_2mg_TiO2_iTRAQ,4992")));
// }
// var diff1 = m.Spectra.Except(n.Spectra).ToList();
// Console.WriteLine(m.Condition.ToString() + " : " + diff1.Count.ToString());
// diff1.ForEach(k =>
// {
// var lst = all.FindAll(l => l.Query.FileScan.LongFileName.Equals(k.Query.FileScan.LongFileName));
// lst.ForEach(q => Console.WriteLine(q.Query.FileScan.ShortFileName + "\t" + q.Tag + "\t" + q.Score.ToString() + "\t" + q.Sequence));
// });
//}
});
}
public void KeepTopPeptideFromSameEngineDifferentParameters()
{
List <IIdentifiedSpectrum> spectra = new List <IIdentifiedSpectrum>(new IIdentifiedSpectrum[] { s1, s2, s4, s5 });
IdentifiedSpectrumUtils.KeepTopPeptideFromSameEngineDifferentParameters(spectra, new ScoreFunction());
Assert.AreEqual(2, spectra.Count);
Assert.IsTrue(spectra.Contains(s1));
Assert.IsTrue(spectra.Contains(s4));
}
public void TestFillProteinInformation()
{
var peptides = new MascotPeptideTextFormat().ReadFromFile(TestContext.CurrentContext.TestDirectory + "/../../../data/Test.output.xml.FDR0.01.peptides");
Assert.IsTrue(peptides.All(m => m.Peptide.Proteins.Count == 0));
IdentifiedSpectrumUtils.FillProteinInformation(peptides, TestContext.CurrentContext.TestDirectory + "/../../../data//Test.output.xml.FDR0.01.peptides.proteins");
Assert.IsTrue(peptides.All(m => m.Peptide.Proteins.Count > 0));
}
public void KeepUnconflictPeptidesFromSameEngineDifferentParameters()
{
List <IIdentifiedSpectrum> spectra = new List <IIdentifiedSpectrum>(new IIdentifiedSpectrum[] { s1, s2, s6, s7 });
IdentifiedSpectrumUtils.KeepUnconflictPeptidesFromSameEngineDifferentParameters(spectra, new ScoreFunction());
Assert.AreEqual(3, spectra.Count);
Assert.IsTrue(spectra.Contains(s1));
Assert.IsTrue(spectra.Contains(s6));
Assert.IsTrue(spectra.Contains(s7));
}
public void TestCalculateQValue()
{
var peptides = new MascotPeptideTextFormat().ReadFromFile(TestContext.CurrentContext.TestDirectory + "/../../../data/QTOF_Ecoli.LowRes.t.xml.peptides");
peptides.RemoveAll(m => m.ExpectValue > 0.05 || m.Peptide.PureSequence.Length < 6);
peptides.ForEach(m => m.FromDecoy = m.Proteins.Any(l => l.Contains("REVERSE_")));
IdentifiedSpectrumUtils.CalculateQValue(peptides, new ExpectValueFunction(), new TargetFalseDiscoveryRateCalculator());
Assert.AreEqual(0.0267, peptides[0].QValue, 0.0001);
}
public void TestRemoveSameSpectrumWithDifferentCharge()
{
s1.Score = 3.0;
s2.Score = 2.0;
s3.Score = 2.0;
s6.Score = 7.0;
List <IIdentifiedSpectrum> spectra = new List <IIdentifiedSpectrum>(new IIdentifiedSpectrum[] { s1, s2, s3, s4, s5, s6 });
IdentifiedSpectrumUtils.FilterSameSpectrumWithDifferentCharge(spectra);
Assert.AreEqual(3, spectra.Count);
Assert.IsTrue(spectra.Contains(s1));
Assert.IsTrue(spectra.Contains(s2));
Assert.IsTrue(spectra.Contains(s6));
s3.Score = 4.0;
s6.Score = 3.0;
spectra = new List <IIdentifiedSpectrum>(new IIdentifiedSpectrum[] { s1, s2, s3, s4, s5, s6 });
IdentifiedSpectrumUtils.FilterSameSpectrumWithDifferentCharge(spectra);
Assert.AreEqual(3, spectra.Count);
Assert.IsTrue(spectra.Contains(s3));
Assert.IsTrue(spectra.Contains(s4));
Assert.IsTrue(spectra.Contains(s5));
}
private void DoInitUniquePeptideCount(IEnumerable <IIdentifiedPeptide> peps)
{
this.uniquePeptideCount = IdentifiedPeptideUtils.GetUniquePeptideCount(peps);
this.peptideCount = IdentifiedSpectrumUtils.GetSpectrumCount(from p in peps select p.Spectrum);
}
public static IdentificationSummary Parse(string proteinFile, string defaultDecoyPattern, IFalseDiscoveryRateCalculator defaultCalc)
{
IdentificationSummary result = new IdentificationSummary();
result.FileName = FileUtils.ChangeExtension(new FileInfo(proteinFile).Name, "");
Regex decoyReg = new Regex(defaultDecoyPattern);
IIdentifiedProteinGroupFilter decoyFilter = null;
IFalseDiscoveryRateCalculator curCalc = null;
var paramFile = FileUtils.ChangeExtension(proteinFile, ".param");
if (File.Exists(paramFile))
{
BuildSummaryOptions options = BuildSummaryOptionsUtils.LoadFromFile(paramFile);
if (options.FalseDiscoveryRate.FilterByFdr)
{
decoyFilter = options.GetDecoyGroupFilter();
curCalc = options.FalseDiscoveryRate.GetFalseDiscoveryRateCalculator();
}
}
if (decoyFilter == null)
{
decoyFilter = new IdentifiedProteinGroupNameRegexFilter(defaultDecoyPattern, false);
curCalc = defaultCalc;
}
var peptideFile = FileUtils.ChangeExtension(proteinFile, ".peptides");
if (File.Exists(peptideFile))
{
var peptides = new MascotPeptideTextFormat().ReadFromFile(peptideFile);
var fullSpectra = GetSpectraByNPT(peptides, 2);
var fullTargetSpectra = GetTargetSpectra(decoyReg, fullSpectra);
var semiSpectra = GetSpectraByNPT(peptides, 1);
var semiTargetSpectra = GetTargetSpectra(decoyReg, semiSpectra);
result.FullSpectrumCount = GetSpectrumCount(fullSpectra);
result.FullTargetSpectrumCount = GetSpectrumCount(fullTargetSpectra);
result.SemiSpectrumCount = GetSpectrumCount(semiSpectra);
result.SemiTargetSpectrumCount = GetSpectrumCount(semiTargetSpectra);
result.FullPeptideCount = IdentifiedSpectrumUtils.GetUniquePeptideCount(fullSpectra);
result.FullTargetPeptideCount = IdentifiedSpectrumUtils.GetUniquePeptideCount(fullTargetSpectra);
result.SemiPeptideCount = IdentifiedSpectrumUtils.GetUniquePeptideCount(semiSpectra);
result.SemiTargetPeptideCount = IdentifiedSpectrumUtils.GetUniquePeptideCount(semiTargetSpectra);
result.FullSpectrumFdr = curCalc.Calculate(result.FullSpectrumCount - result.FullTargetSpectrumCount, result.FullTargetSpectrumCount);
result.SemiSpectrumFdr = curCalc.Calculate(result.SemiSpectrumCount - result.SemiTargetSpectrumCount, result.SemiTargetSpectrumCount);
result.FullPeptideFdr = curCalc.Calculate(result.FullPeptideCount - result.FullTargetPeptideCount, result.FullTargetPeptideCount);
result.SemiPeptideFdr = curCalc.Calculate(result.SemiPeptideCount - result.SemiTargetPeptideCount, result.SemiTargetPeptideCount);
}
if (File.Exists(proteinFile))
{
var ir = new MascotResultTextFormat().ReadFromFile(proteinFile);
ir.InitUniquePeptideCount();
var u2proteins = (from p in ir
where p[0].UniquePeptideCount > 1
select p).ToList();
var u1proteins = (from p in ir
where p[0].UniquePeptideCount == 1
select p).ToList();
result.ProteinGroupCount = ir.Count;
result.Unique2ProteinGroupCount = u2proteins.Count;
int targetCount;
result.Unique2ProteinFdr = CalculateProteinFdr(u2proteins, decoyFilter, defaultCalc, out targetCount);
result.Unique2ProteinGroupTargetCount = (int)targetCount;
result.Unique1ProteinFdr = CalculateProteinFdr(u1proteins, decoyFilter, defaultCalc, out targetCount);
result.Unique1ProteinGroupTargetCount = (int)targetCount;
}
return(result);
}
private List <IIdentifiedSpectrum> DoCalculateDuplicated(List <IIdentifiedSpectrum> preFiltered, OptimalResult optimalResult)
{
preFiltered.ForEach(m => m.QValue = -1);
ScoreFunc.SortSpectrum(preFiltered);
//Use top score spectra for Qvalue calculation
var topSpectra = new List <IIdentifiedSpectrum>(preFiltered);
IdentifiedSpectrumUtils.KeepTopPeptideFromSameEngineDifferentParameters(topSpectra, ScoreFunc);
//计算QValue。
QValueFunc(topSpectra, ScoreFunc, FdrCalc);
//将非top的肽段的QValue填充。
for (int i = 1; i < preFiltered.Count; i++)
{
if (preFiltered[i].QValue == -1)
{
preFiltered[i].QValue = preFiltered[i - 1].QValue;
}
}
List <IIdentifiedSpectrum> result = new List <IIdentifiedSpectrum>();
optimalResult.PeptideCountFromDecoyDB = 0;
optimalResult.PeptideCountFromTargetDB = 0;
for (int i = preFiltered.Count - 1; i >= 0; i--)
{
if (preFiltered[i].QValue <= FdrValue)
{
result.AddRange(preFiltered.GetRange(0, i + 1));
optimalResult.Score = ScoreFunc.GetScore(preFiltered[i]);
optimalResult.ExpectValue = preFiltered[i].ExpectValue;
optimalResult.FalseDiscoveryRate = preFiltered[i].QValue;
int decoyCount = 0;
int targetCount = 0;
HashSet <string> filenames = new HashSet <string>();
foreach (IIdentifiedSpectrum spectrum in result)
{
if (filenames.Contains(spectrum.Query.FileScan.LongFileName))
{
continue;
}
filenames.Add(spectrum.Query.FileScan.LongFileName);
if (spectrum.FromDecoy)
{
decoyCount++;
}
else
{
targetCount++;
}
}
optimalResult.PeptideCountFromDecoyDB = decoyCount;
optimalResult.PeptideCountFromTargetDB = targetCount;
Console.WriteLine("{0} -> {1} ==> {2} / {3}", preFiltered.Count, topSpectra.Count, decoyCount, targetCount);
break;
}
}
return(result);
}
public void CalculateQValue(List <IIdentifiedSpectrum> spectra)
{
IdentifiedSpectrumUtils.CalculateQValue(spectra, scoreFunc, fdrCalc);
}