public void TestFeatureAlignment()
{
const string outFilePath = @"\\protoapps\UserData\Jungkap\Lewy\aligned\promex_crosstab_temp.tsv";
//CPTAC_Intact_CR32A_24Aug15_Bane_15-02-06-RZ
var prsmReader = new ProteinSpectrumMatchReader();
var tolerance = new Tolerance(10);
var alignment = new LcMsFeatureAlignment(new AnalysisCompRef.CompRefFeatureComparer(tolerance));
for (var i = 0; i < NdataSet; i++)
{
var rawFile = string.Format(@"{0}\{1}.pbf", PbfPath, GetDataSetNames(i));
var mspFile = string.Format(@"{0}\{1}_IcTda.tsv", MsPfFolder, GetDataSetNames(i));
var mspFile2 = string.Format(@"{0}\{1}_IcTda.tsv", MsPfFolder2, GetDataSetNames(i));
var ms1FtFile = string.Format(@"{0}\{1}.ms1ft", Ms1FtFolder, GetDataSetNames(i));
Console.WriteLine(rawFile);
var run = PbfLcMsRun.GetLcMsRun(rawFile);
var prsmList1 = prsmReader.LoadIdentificationResult(mspFile, ProteinSpectrumMatch.SearchTool.MsPathFinder);
var prsmList2 = prsmReader.LoadIdentificationResult(mspFile2, ProteinSpectrumMatch.SearchTool.MsPathFinder);
prsmList1.AddRange(prsmList2);
var prsmList = MergePrsm(prsmList1);
var features = LcMsFeatureAlignment.LoadProMexResult(i, ms1FtFile, run);
for (var j = 0; j < prsmList.Count; j++)
{
var match = prsmList[j];
match.ProteinId = match.ProteinName;
}
// tag features by PrSMs
for (var j = 0; j < features.Count; j++)
{
//features[j].ProteinSpectrumMatches = new ProteinSpectrumMatchSet(i);
var massTol = tolerance.GetToleranceAsTh(features[j].Mass);
foreach (var match in prsmList)
{
if (features[j].MinScanNum < match.ScanNum && match.ScanNum < features[j].MaxScanNum && Math.Abs(features[j].Mass - match.Mass) < massTol)
{
features[j].ProteinSpectrumMatches.Add(match);
}
}
}
alignment.AddDataSet(i, features, run);
}
alignment.AlignFeatures();
Console.WriteLine("{0} alignments ", alignment.CountAlignedFeatures);
for (var i = 0; i < NdataSet; i++)
{
alignment.FillMissingFeatures(i);
Console.WriteLine("{0} has been processed", GetDataSetNames(i));
}
OutputCrossTabWithId(outFilePath, alignment);
}
public static ICollection <ProteinSpectrumMatchSet> CollectTrainSet(string pbfFilePath, string idFilePath)
{
Modification.RegisterAndGetModification(Modification.Cysteinyl.Name, Modification.Cysteinyl.Composition);
var prsmReader = new ProteinSpectrumMatchReader(0.01);
var prsmList = prsmReader.LoadIdentificationResult(idFilePath);
var run = PbfLcMsRun.GetLcMsRun(pbfFilePath);
var groupedPrsmList = GroupingByPrsm(0, prsmList, new PrsmComparer(run));
var finalPrsmGroups = new List <ProteinSpectrumMatchSet>();
foreach (var prsmSet in groupedPrsmList)
{
if (prsmSet.Count < 2)
{
continue;
}
var isGood = false;
var sequence = prsmSet[0].GetSequence();
if (sequence == null)
{
continue;
}
foreach (var scan in prsmSet.Select(prsm => prsm.ScanNum))
{
var spectrum = run.GetSpectrum(scan) as ProductSpectrum;
if (spectrum == null)
{
continue;
}
if (IsGoodTarget(spectrum, sequence))
{
isGood = true;
break;
}
}
if (isGood)
{
finalPrsmGroups.Add(prsmSet);
}
}
return(finalPrsmGroups);
}
public void TestGenerateFrequencyData()
{
var methodName = MethodBase.GetCurrentMethod().Name;
TestUtils.ShowStarting(methodName);
const string idFileFolder = @"D:\MassSpecFiles\training\IdScoring\MSPF_trainset";
const string outFileFolder = @"D:\MassSpecFiles\training\IdScoring";
if (!Directory.Exists(idFileFolder))
{
Assert.Ignore(@"Skipping test {0} since folder not found: {1}", methodName, idFileFolder);
}
Modification.RegisterAndGetModification(Modification.Cysteinyl.Name, Modification.Cysteinyl.Composition);
Modification.RegisterAndGetModification(Modification.Phosphorylation.Name, Modification.Phosphorylation.Composition);
Modification.RegisterAndGetModification(Modification.Methylation.Name, Modification.Methylation.Composition);
Modification.RegisterAndGetModification(Modification.DiMethylation.Name, Modification.DiMethylation.Composition);
Modification.RegisterAndGetModification(Modification.TriMethylation.Name, Modification.TriMethylation.Composition);
Modification.RegisterAndGetModification("Trioxidation", new Composition(0, 0, 0, 3, 0));
var aaSet = new AminoAcidSet(@"D:\MassSpecFiles\training\Mods.txt");
var n = 0;
for (var d = 0; d < TrainSetFileLists.Length; d++)
{
var dataset = TrainSetFileLists[d];
var dataname = Path.GetFileNameWithoutExtension(dataset);
var idFile = string.Format(@"{0}\{1}_IcTda.tsv", idFileFolder, dataname);
var decoyFile = string.Format(@"{0}\{1}_IcDecoy.tsv", idFileFolder, dataname);
var targetFile = string.Format(@"{0}\{1}_IcTarget.tsv", idFileFolder, dataname);
if (!File.Exists(idFile)) continue;
var prsmReader = new ProteinSpectrumMatchReader(0.01);
var prsmList = prsmReader.LoadIdentificationResult(idFile);
var minScore = prsmList.Last().Score;
var decoyMatches = prsmReader.ReadMsPathFinderResult(decoyFile, int.MaxValue, 1, Math.Max(minScore - 5, 10));
var run = PbfLcMsRun.GetLcMsRun(dataset);
var spectrumMatchSet = LcMsFeatureTrain.CollectTrainSet(dataset, idFile);
Console.WriteLine(spectrumMatchSet.Count);
var writer = new StreamWriter(string.Format(@"{0}\{1}_target.tsv", outFileFolder, dataname));
foreach (var matches in spectrumMatchSet)
{
foreach (var match in matches)
{
var spec = run.GetSpectrum(match.ScanNum) as ProductSpectrum;
GetMatchStatistics(spec, match.GetSequence(), match.Charge, writer);
}
}
writer.Close();
writer = new StreamWriter(string.Format(@"{0}\{1}_decoy.tsv", outFileFolder, dataname));
foreach (var match in decoyMatches)
{
var sequence = match.GetSequence();
var spec = run.GetSpectrum(match.ScanNum) as ProductSpectrum;
GetMatchStatistics(spec, sequence, match.Charge, writer);
}
writer.Close();
n++;
}
}
public void FindMissingLcMsFeatures()
{
var mspfFolder = @"D:\MassSpecFiles\CompRef_Kelleher\Study3";
var ms1ftFolder = @"D:\MassSpecFiles\CompRef_Kelleher\Study3";
const int Nfraction1 = 3;
const int Nfraction2 = 5;
for (var frac1 = 1; frac1 <= Nfraction1; frac1++)
{
for (var frac2 = 1; frac2 <= Nfraction2; frac2++)
{
var datasets = GetDataSetNamesStudy3(frac1, frac2);
//var outFilePath = string.Format(@"D:\MassSpecFiles\CompRef_Kelleher\study3_GFrep{0}_Gfrac{1}.tsv", frac1.ToString("D2"), frac2.ToString("D2"));
var nDataset = datasets.Count;
var prsmReader = new ProteinSpectrumMatchReader();
var tolerance = new Tolerance(12);
for (var i = 0; i < nDataset; i++)
{
var rawFile = string.Format(@"{0}\{1}.pbf", PbfPath, datasets[i]);
var mspFile = string.Format(@"{0}\{1}_IcTda.tsv", mspfFolder, datasets[i]);
var ms1FtFile = string.Format(@"{0}\{1}.ms1ft", ms1ftFolder, datasets[i]);
var outPath = string.Format(@"{0}\{1}.seqtag.ms1ft", ms1ftFolder, datasets[i]);
if (File.Exists(outPath)) continue;
var run = PbfLcMsRun.GetLcMsRun(rawFile);
var features = LcMsFeatureAlignment.LoadProMexResult(i, ms1FtFile, run);
var prsmList = prsmReader.LoadIdentificationResult(mspFile, ProteinSpectrumMatch.SearchTool.MsPathFinder);
var prsmFeatureMatch = new bool[prsmList.Count];
for (var j = 0; j < features.Count; j++)
{
//features[j].ProteinSpectrumMatches = new ProteinSpectrumMatchSet(i);
var massTol = tolerance.GetToleranceAsTh(features[j].Mass);
for (var k = 0; k < prsmList.Count; k++)
{
var match = prsmList[k];
if (features[j].MinScanNum < match.ScanNum && match.ScanNum < features[j].MaxScanNum && Math.Abs(features[j].Mass - match.Mass) < massTol)
{
features[j].ProteinSpectrumMatches.Add(match);
prsmFeatureMatch[k] = true;
}
}
}
var missingPrsm = new List<ProteinSpectrumMatch>();
for (var k = 0; k < prsmList.Count; k++) if (!prsmFeatureMatch[k]) missingPrsm.Add(prsmList[k]);
FeatureFind(missingPrsm, run, outPath);
Console.WriteLine(outPath);
}
}
}
}
public void AlignFeatures(List<string> datasets, string mspfFolder, string ms1ftFolder, string outFilePath)
{
var nDataset = datasets.Count;
var prsmReader = new ProteinSpectrumMatchReader();
var tolerance = new Tolerance(12);
var alignment = new LcMsFeatureAlignment(new AnalysisCompRef.CompRefFeatureComparer(tolerance));
for (var i = 0; i < nDataset; i++)
{
var rawFile = string.Format(@"{0}\{1}.pbf", PbfPath, datasets[i]);
var mspFile = string.Format(@"{0}\{1}_IcTda.tsv", mspfFolder, datasets[i]);
var ms1FtFile = string.Format(@"{0}\{1}.ms1ft", ms1ftFolder, datasets[i]);
var ms1FtFile2 = string.Format(@"{0}\{1}.seqtag.ms1ft", ms1ftFolder, datasets[i]);
var run = PbfLcMsRun.GetLcMsRun(rawFile);
var features = LcMsFeatureAlignment.LoadProMexResult(i, ms1FtFile, run);
var features2 = LcMsFeatureAlignment.LoadProMexResult(i, ms1FtFile2, run);
features.AddRange(features2);
if (File.Exists(mspFile))
{
var prsmList = prsmReader.LoadIdentificationResult(mspFile, ProteinSpectrumMatch.SearchTool.MsPathFinder);
//var prsmFeatureMatch = new bool[prsmList.Count];
for (var j = 0; j < prsmList.Count; j++)
{
var match = prsmList[j];
match.ProteinId = match.ProteinName;
}
// tag features by PrSMs
for (var j = 0; j < features.Count; j++)
{
//features[j].ProteinSpectrumMatches = new ProteinSpectrumMatchSet(i);
var massTol = tolerance.GetToleranceAsTh(features[j].Mass);
for(var k = 0; k < prsmList.Count; k++)
{
var match = prsmList[k];
if (features[j].MinScanNum < match.ScanNum && match.ScanNum < features[j].MaxScanNum && Math.Abs(features[j].Mass - match.Mass) < massTol)
{
features[j].ProteinSpectrumMatches.Add(match);
//prsmFeatureMatch[k] = true;
}
}
}
}
alignment.AddDataSet(i, features, run);
}
alignment.AlignFeatures();
Console.WriteLine("{0} alignments ", alignment.CountAlignedFeatures);
for (var i = 0; i < nDataset; i++)
{
alignment.FillMissingFeatures(i);
Console.WriteLine("{0} has been processed", datasets[i]);
}
AnalysisCompRef.OutputCrossTabWithId(outFilePath, alignment, datasets.ToArray());
}
