private void Rescore(string msAlignFilePath, string outputFilePath)
{
var parser = new TsvFileParser(msAlignFilePath);
var sequences = parser.GetData("Peptide");
var scanNums = parser.GetData("Scan(s)").Select(s => Convert.ToInt32(s)).ToArray();
var charges = parser.GetData("Charge").Select(c => Convert.ToInt32(c)).ToArray();
var rows = parser.GetRows();
var headers = parser.GetHeaders();
using (var writer = new StreamWriter(outputFilePath))
{
writer.WriteLine("{0}\t{1}", string.Join("\t", headers), IcScores.GetScoreNames());
for (var i = 0; i < parser.NumData; i++)
{
var row = rows[i];
var seqStr = SimpleStringProcessing.GetStringBetweenDots(sequences[i]);
if (seqStr == null || seqStr.Contains("("))
{
continue; //TODO: currently ignore ids with modifications
}
var composition = AASet.GetComposition(seqStr);
//var sequence = new Sequence(seqStr, AASet);
//if (sequence == null)
//{
// Console.WriteLine("Ignore illegal sequence: {0}", seqStr);
// continue;
//}
var charge = charges[i];
var scanNum = scanNums[i];
var scores = _topDownScorer.GetScores(AminoAcid.ProteinNTerm, seqStr, AminoAcid.ProteinCTerm, composition, charge, scanNum);
if (scores == null)
{
continue;
}
writer.WriteLine("{0}\t{1}", row, scores);
}
}
}
private void Rescore(string icResultFilePath, string outputFilePath)
{
var parser = new TsvFileParser(icResultFilePath);
var sequences = parser.GetData("Sequence");
var scanNums = parser.GetData("ScanNum").Select(s => Convert.ToInt32(s)).ToArray();
var charges = parser.GetData("Charge").Select(c => Convert.ToInt32(c)).ToArray();
var compositions = parser.GetData("Composition").Select(Composition.Parse).ToArray();
var modIndex = parser.GetHeaders().IndexOf("Modifications");
var rows = parser.GetRows();
var headers = parser.GetHeaders();
using (var writer = new StreamWriter(outputFilePath))
{
writer.WriteLine("{0}\t{1}", string.Join("\t", headers), IcScores.GetScoreNames());
for (var i = 0; i < parser.NumData; i++)
{
var row = rows[i];
var seqStr = sequences[i];
var charge = charges[i];
var scanNum = scanNums[i];
var composition = compositions[i];
var scores = _topDownScorer.GetScores(AminoAcid.ProteinNTerm, seqStr, AminoAcid.ProteinCTerm, composition, charge, scanNum);
var token = row.Split('\t');
for (var j = 0; j < token.Length; j++)
{
if (j != modIndex)
{
writer.Write(token[j] + "\t");
}
else
{
writer.Write("[" + scores.Modifications + "]" + "\t");
}
}
writer.WriteLine(scores);
}
}
}
public void TestTagAlignedFeatures()
{
var methodName = MethodBase.GetCurrentMethod().Name;
Utils.ShowStarting(methodName);
var featureDir = Path.Combine(Utils.DEFAULT_TEST_FILE_FOLDER, "Output");
var mspDir = Path.Combine(Utils.DEFAULT_TEST_FILE_FOLDER, @"Output\MSP");
var outFile = Path.Combine(Utils.DEFAULT_TEST_FILE_FOLDER, @"Output\aligned_features.tsv");
var resultFile = Path.Combine(Utils.DEFAULT_TEST_FILE_FOLDER, @"\Output\aligned_ids.tsv");
if (!Directory.Exists(featureDir))
{
Assert.Ignore(@"Skipping test {0} since folder not found: {1}", methodName, featureDir);
}
if (!Directory.Exists(mspDir))
{
Assert.Ignore(@"Skipping test {0} since folder not found: {1}", methodName, mspDir);
}
if (!File.Exists(outFile))
{
Assert.Ignore(@"Skipping test {0} since file not found: {1}", methodName, outFile);
}
var dataset = GetDataList(featureDir);
var tsvParser = new TsvFileParser(outFile);
var massList = new List <double>();
for (var i = 0; i < tsvParser.NumData; i++)
{
massList.Add(Double.Parse(tsvParser.GetData("MonoMass")[i]));
}
var featureIdMap = new Dictionary <int, string>();
var tolerance = new Tolerance(12);
var headers = new List <string>();
//foreach (var data in dataset)
for (var d = 0; d < dataset.Count; d++)
{
var data = dataset[d];
var minScanColName = string.Format("{0}_minScan", d);
var maxScanColName = string.Format("{0}_maxScan", d);
var fname = string.Format(@"{0}\{1}_IcTda.tsv", mspDir, data);
var idParser = new TsvFileParser(fname);
var idRows = idParser.GetRows();
if (headers.Count < 1)
{
headers.AddRange(idParser.GetHeaders());
}
for (var i = 0; i < idParser.NumData; i++)
{
var scan = Int32.Parse(idParser.GetData("Scan")[i]);
var mass = Double.Parse(idParser.GetData("Mass")[i]);
var qvalue = Double.Parse(idParser.GetData("QValue")[i]);
if (qvalue > 0.01)
{
break;
}
var massTol = tolerance.GetToleranceAsMz(mass);
var idx = massList.BinarySearch(mass);
if (idx < 0)
{
idx = ~idx;
}
var found = false;
for (var j = idx; j >= 0; j--)
{
if (Math.Abs(mass - massList[j]) > massTol)
{
break;
}
if (tsvParser.GetData(minScanColName)[j].Length < 1)
{
continue;
}
if (Int32.Parse(tsvParser.GetData(minScanColName)[j]) < scan && scan < Int32.Parse(tsvParser.GetData(maxScanColName)[j]))
{
found = true;
if (!featureIdMap.ContainsKey(j))
{
featureIdMap.Add(j, idRows[i]);
}
break;
}
}
if (found)
{
continue;
}
for (var j = idx + 1; j < massList.Count; j++)
{
if (Math.Abs(mass - massList[j]) > massTol)
{
break;
}
if (tsvParser.GetData(minScanColName)[j].Length < 1)
{
continue;
}
if (Int32.Parse(tsvParser.GetData(minScanColName)[j]) < scan && scan < Int32.Parse(tsvParser.GetData(maxScanColName)[j]))
{
found = true;
if (!featureIdMap.ContainsKey(j))
{
featureIdMap.Add(j, idRows[i]);
}
break;
}
}
}
}
var writer = new StreamWriter(resultFile);
writer.Write("AlignedFeatureID"); writer.Write("\t");
writer.Write(string.Join("\t", headers));
for (var i = 0; i < 32; i++)
{
writer.Write("\t"); writer.Write("{0}", i);
}
writer.Write("\n");
var id = 1;
foreach (var key in featureIdMap.Keys)
{
writer.Write(id); writer.Write("\t");
writer.Write(featureIdMap[key]);
for (var i = 0; i < 32; i++)
{
writer.Write("\t"); writer.Write("{0}", tsvParser.GetData(string.Format("{0}", i))[key]);
}
writer.Write("\n");
id++;
}
writer.Close();
}
public void TestCompositeScoring()
{
var methodName = MethodBase.GetCurrentMethod().Name;
TestUtils.ShowStarting(methodName);
//const string rawFilePath = @"\\proto-2\UnitTest_Files\InformedProteomics_TestFiles\SpecFiles\QC_Shew_Intact_26Sep14_Bane_C2Column3.raw";
const string rawFilePath = @"D:\MassSpecFiles\training\raw\QC_Shew_Intact_26Sep14_Bane_C2Column3.pbf";
if (!File.Exists(rawFilePath))
{
Assert.Ignore(@"Skipping test {0} since file not found: {1}", methodName, rawFilePath);
}
// Configure amino acid set
var oxM = new SearchModification(Modification.Oxidation, 'M', SequenceLocation.Everywhere, false);
var dehydroC = new SearchModification(Modification.Dehydro, 'C', SequenceLocation.Everywhere, false);
var acetylN = new SearchModification(Modification.Acetylation, '*', SequenceLocation.ProteinNTerm, false);
const int numMaxModsPerProtein = 4;
var searchModifications = new List <SearchModification>
{
dehydroC,
oxM,
acetylN
};
var aaSet = new AminoAcidSet(searchModifications, numMaxModsPerProtein);
var comparer = new FilteredProteinMassBinning(aaSet, 50000, 28);
var run = PbfLcMsRun.GetLcMsRun(rawFilePath);
const double filteringWindowSize = 1.1;
const int isotopeOffsetTolerance = 2;
var tolerance = new Tolerance(10);
const int minCharge = 1;
const int maxCharge = 20;
var graphFactory = new ProteinScoringGraphFactory(comparer, aaSet);
var aminoAcidSet = new AminoAcidSet();
//var scorer = new MatchedPeakPostScorer(tolerance, minCharge, maxCharge);
var scorer = new InformedTopDownScorer(run, aminoAcidSet, minCharge, maxCharge, tolerance);
var fileExt = new string[] { "IcTarget", "IcDecoy" };
foreach (var ext in fileExt)
{
var resultFileName = string.Format(@"D:\MassSpecFiles\training\Rescoring\QC_Shew_Intact_26Sep14_Bane_C2Column3_{0}.tsv", ext);
var parser = new TsvFileParser(resultFileName);
var scans = parser.GetData("Scan").Select(s => Convert.ToInt32(s)).ToArray();
var charges = parser.GetData("Charge").Select(s => Convert.ToInt32(s)).ToArray();
var protSequences = parser.GetData("Sequence").ToArray();
var modStrs = parser.GetData("Modifications").ToArray();
var compositions = parser.GetData("Composition").Select(Composition.Parse).ToArray();
var protMass = parser.GetData("Mass").Select(s => Convert.ToDouble(s)).ToArray();
var outputFileName = string.Format(@"D:\MassSpecFiles\training\Rescoring\QC_Shew_Intact_26Sep14_Bane_C2Column3_{0}_Rescored.tsv", ext);
using (var writer = new StreamWriter(outputFileName))
{
writer.WriteLine(string.Join("\t", parser.GetHeaders().ToArray(), 0, 15) + "\tScore\tEValue");
var lines = new string[parser.NumData];
//for (var i = 0; i < parser.NumData; i++)
Parallel.For(0, parser.NumData, i =>
{
var scan = scans[i];
var charge = charges[i];
var protSequence = protSequences[i];
var modStr = modStrs[i];
var sequence = Sequence.CreateSequence(protSequence, modStr, aminoAcidSet);
Assert.True(sequence.Composition.Equals(compositions[i] - Composition.H2O));
var ms2Spec = run.GetSpectrum(scan) as ProductSpectrum;
Assert.True(ms2Spec != null);
var scores = scorer.GetScores(sequence, charge, scan);
var deconvSpec = Deconvoluter.GetDeconvolutedSpectrum(ms2Spec, minCharge, maxCharge,
isotopeOffsetTolerance, filteringWindowSize, tolerance, 0.7);
var deconvScorer = new CompositeScorerBasedOnDeconvolutedSpectrum(deconvSpec, ms2Spec, tolerance,
comparer);
var graph = graphFactory.CreateScoringGraph(deconvScorer, protMass[i]);
var gf = new GeneratingFunction(graph);
gf.ComputeGeneratingFunction();
var specEvalue = gf.GetSpectralEValue(scores.Score);
var rowStr = parser.GetRows()[i];
var items = rowStr.Split('\t').ToArray();
var newRowStr = string.Join("\t", items, 0, 15);
//writer.WriteLine("{0}\t{1}\t{2}", newRowStr, scores.Score, specEvalue);
lock (lines)
{
lines[i] = string.Format("{0}\t{1}\t{2}", newRowStr, scores.Score, specEvalue);
}
//Console.WriteLine("{0}\t{1}\t{2}", items[0], scores.Score, specEvalue);
});
foreach (var line in lines)
{
writer.WriteLine(line);
}
}
Console.WriteLine("Done");
}
}
public void CreatePeptideAbundanceTableWithSkyline()
{
var methodName = MethodBase.GetCurrentMethod().Name;
Utils.ShowStarting(methodName);
// Reading Henry's results
var pepKeySet = new HashSet <string>();
var resultDic = new Dictionary <string, Tuple <double, double> >();
const string henryResultPath = @"H:\Research\IPRG2015\Henry_results\tsv";
if (!Directory.Exists(henryResultPath))
{
Assert.Ignore(@"Skipping test {0} since folder not found: {1}", methodName, henryResultPath);
}
var aaSet = new AminoAcidSet();
foreach (var resultFile in Directory.GetFiles(henryResultPath, "*.tsv"))
{
var fileName = Path.GetFileName(resultFile);
if (fileName == null)
{
continue;
}
var sample = fileName.Substring(0, 2);
Console.WriteLine("Processing {0}", sample);
var tsvReader = new TsvFileParser(resultFile);
var peptides = tsvReader.GetData("Peptide").ToArray();
var charge = tsvReader.GetData("Charge").Select(c => Convert.ToInt32(c)).ToArray();
var prob = tsvReader.GetData("Prob").Select(Convert.ToDouble).ToArray();
var qValue = tsvReader.GetData("QValue").Select(Convert.ToDouble).ToArray();
for (var i = 0; i < tsvReader.NumData; i++)
{
var peptide = peptides[i];
var nominalMass = GetNominalMass(aaSet, peptide);
var key = sample + ":" + GetPeptide(peptides[i]) + ":" + nominalMass + ":" + charge[i];
var pepKey = GetPeptide(peptides[i]) + ":" + nominalMass;
pepKeySet.Add(pepKey);
Tuple <double, double> existingScores;
if (resultDic.TryGetValue(key, out existingScores))
{
if (prob[i] > existingScores.Item1)
{
resultDic[key] = new Tuple <double, double>(prob[i], qValue[i]);
}
}
else
{
resultDic.Add(key, new Tuple <double, double>(prob[i], qValue[i]));
}
}
}
const string skylineFilePath = @"H:\Research\IPRG2015\MySkyline\TransitionResults.csv";
if (!File.Exists(skylineFilePath))
{
Assert.Ignore(@"Skipping test {0} since file not found: {1}", methodName, skylineFilePath);
}
var skylineTable = new TsvFileParser(skylineFilePath, ',');
const string outputFilePath = @"H:\Research\IPRG2015\MySkyline\SkylineTransitionResultsWithScores3.tsv";
using (var writer = new StreamWriter(outputFilePath))
{
var peptides = skylineTable.GetData("Peptide Sequence").ToArray();
var samples = skylineTable.GetData("Replicate Name").Select(s => "" + s[0] + s[2]).ToArray();
var charges = skylineTable.GetData("Precursor Charge").Select(c => Convert.ToInt32(c)).ToArray();
var precursorMzs = skylineTable.GetData("Precursor Mz").Select(Convert.ToDouble).ToArray();
writer.WriteLine("{0}\tProbability\tQValue", string.Join("\t", skylineTable.GetHeaders().Take(skylineTable.GetHeaders().Count - 2)));
for (var i = 0; i < skylineTable.NumData; i++)
{
var precursorMz = precursorMzs[i];
var charge = charges[i];
var nominalMass = (int)Math.Round(((precursorMz - Constants.Proton) * charge - Composition.H2O.Mass) *
Constants.RescalingConstant);
var pepKey = peptides[i] + ":" + nominalMass;
if (!pepKeySet.Contains(pepKey))
{
//Console.WriteLine("Removing {0}", pepKey);
continue;
}
var key = samples[i] + ":" + peptides[i] + ":" + nominalMass + ":" + charge;
double?prob = null, qValue = null;
Tuple <double, double> scores;
if (resultDic.TryGetValue(key, out scores))
{
prob = scores.Item1;
qValue = scores.Item2;
}
var skylineData = skylineTable.GetRows()[i].Split(',');
for (var j = 0; j < skylineData.Length - 2; j++)
{
if (j != 2)
{
writer.Write(skylineData[j] + "\t");
}
else
{
writer.Write("" + skylineData[j][0] + skylineData[j][2] + "\t");
}
}
writer.WriteLine("{0}\t{1}",
prob != null ? prob.ToString() : "NA",
qValue != null ? qValue.ToString() : "NA");
}
}
Console.WriteLine("Done");
}
public void AddMostAbundantIsotopePeakIntensity()
{
var methodName = MethodBase.GetCurrentMethod().Name;
Utils.ShowStarting(methodName);
var rawFilePath = Path.Combine(Utils.DEFAULT_TEST_FILE_FOLDER, @"TestYufengData\QC_ShewIntact_40K_LongSeparation_1_141016155143.raw");
if (!File.Exists(rawFilePath))
{
Assert.Ignore(@"Skipping test " + methodName + @" since file not found: " + rawFilePath);
}
var run = PbfLcMsRun.GetLcMsRun(rawFilePath);
var resultFilePath = Path.Combine(Utils.DEFAULT_TEST_FILE_FOLDER, @"TestYufengData\QC_ShewIntact_40K_LongSeparation_1_141016155143_IcTda.tsv");
var parser = new TsvFileParser(resultFilePath);
var compositions = parser.GetData("Composition").Select(Composition.Parse).ToArray();
var scanNums = parser.GetData("Scan").Select(s => Convert.ToInt32(s)).ToArray();
var charges = parser.GetData("Charge").Select(s => Convert.ToInt32(s)).ToArray();
var precursorIntensities = new double[parser.NumData];
var tolerance = new Tolerance(10);
for (var i = 0; i < parser.NumData; i++)
{
var scanNum = scanNums[i];
var composition = compositions[i];
var charge = charges[i];
var precursorIon = new Ion(composition, charge);
var precursorScanNum = run.GetPrecursorScanNum(scanNum);
var precursorSpec = run.GetSpectrum(precursorScanNum);
var isotopePeaks = precursorSpec.GetAllIsotopePeaks(precursorIon, tolerance, 0.1);
if (isotopePeaks != null)
{
var maxIntensity = 0.0;
for (var j = 0; j < isotopePeaks.Length; j++)
{
if (isotopePeaks[j] != null && isotopePeaks[j].Intensity > maxIntensity)
{
maxIntensity = isotopePeaks[j].Intensity;
}
}
precursorIntensities[i] = maxIntensity;
}
}
// Writing
var newResultFilePath = Path.Combine(Utils.DEFAULT_TEST_FILE_FOLDER, @"TestYufengData\QC_ShewIntact_40K_LongSeparation_1_141016155143_IcTdaWithIntensities.tsv");
using (var writer = new StreamWriter(newResultFilePath))
{
writer.WriteLine(string.Join("\t", parser.GetHeaders()) + "\t" + "PrecursorIntensity");
for (var i = 0; i < parser.NumData; i++)
{
writer.WriteLine(parser.GetRows()[i] + "\t" + precursorIntensities[i]);
}
}
Console.WriteLine("Done");
}
public void TestCompositeScoring()
{
var methodName = MethodBase.GetCurrentMethod().Name;
Utils.ShowStarting(methodName);
var pbfFilePath = Utils.GetPbfTestFilePath(false);
var pbfFile = Utils.GetTestFile(methodName, pbfFilePath);
// Configure amino acid set
var oxM = new SearchModification(Modification.Oxidation, 'M', SequenceLocation.Everywhere, false);
var dehydroC = new SearchModification(Modification.Dehydro, 'C', SequenceLocation.Everywhere, false);
var acetylN = new SearchModification(Modification.Acetylation, '*', SequenceLocation.ProteinNTerm, false);
const int numMaxModsPerProtein = 4;
var searchModifications = new List <SearchModification>
{
dehydroC,
oxM,
acetylN
};
var aaSet = new AminoAcidSet(searchModifications, numMaxModsPerProtein);
var comparer = new FilteredProteinMassBinning(aaSet, 50000, 28);
var run = PbfLcMsRun.GetLcMsRun(pbfFile.FullName);
const double filteringWindowSize = 1.1;
const int isotopeOffsetTolerance = 2;
var tolerance = new Tolerance(10);
const int minCharge = 1;
const int maxCharge = 20;
var graphFactory = new ProteinScoringGraphFactory(comparer, aaSet);
var aminoAcidSet = new AminoAcidSet();
//var scorer = new MatchedPeakPostScorer(tolerance, minCharge, maxCharge);
var scorer = new InformedTopDownScorer(run, aminoAcidSet, minCharge, maxCharge, tolerance);
if (pbfFile.DirectoryName == null)
{
Assert.Ignore("Ignoring test since cannot determine the parent directory of " + pbfFile.FullName);
}
var fileExt = new string[] { "IcTarget", "IcDecoy" };
foreach (var ext in fileExt)
{
var resultFileName = Path.Combine(pbfFile.DirectoryName, Path.GetFileNameWithoutExtension(pbfFile.Name)) + string.Format("_{0}.tsv", ext);
var parser = new TsvFileParser(resultFileName);
var scans = parser.GetData("Scan").Select(s => Convert.ToInt32((string)s)).ToArray();
var charges = parser.GetData("Charge").Select(s => Convert.ToInt32(s)).ToArray();
var protSequences = parser.GetData("Sequence").ToArray();
var modStrs = parser.GetData("Modifications").ToArray();
var compositions = parser.GetData("Composition").Select(Composition.Parse).ToArray();
var protMass = parser.GetData("Mass").Select(s => Convert.ToDouble(s)).ToArray();
var outputFileName = Path.Combine(pbfFile.DirectoryName, Path.GetFileNameWithoutExtension(pbfFile.Name)) + string.Format("_{0}_Rescored.tsv", ext);
using (var writer = new StreamWriter(outputFileName))
{
writer.WriteLine(string.Join("\t", parser.GetHeaders().ToArray(), 0, 15) + "\tScore\tEValue");
var lines = new string[parser.NumData];
//for (var i = 0; i < parser.NumData; i++)
Parallel.For(0, 30, i =>
{
var scan = scans[i];
var charge = charges[i];
var protSequence = protSequences[i];
var modStr = modStrs[i];
var sequence = Sequence.CreateSequence(protSequence, modStr, aminoAcidSet);
// Assert.True(sequence.Composition.Equals(compositions[i] - Composition.H2O));
var ms2Spec = run.GetSpectrum(scan) as ProductSpectrum;
if (ms2Spec == null)
{
Console.WriteLine("Could not get the spectrum datafor scan {0}", scan);
}
else
{
Assert.True(ms2Spec != null);
var scores = scorer.GetScores(sequence, charge, scan);
var deconvSpec = Deconvoluter.GetDeconvolutedSpectrum(ms2Spec, minCharge, maxCharge,
isotopeOffsetTolerance, filteringWindowSize, tolerance, 0.7);
var deconvScorer = new CompositeScorerBasedOnDeconvolutedSpectrum(deconvSpec, ms2Spec, tolerance,
comparer);
var graph = graphFactory.CreateScoringGraph(deconvScorer, protMass[i]);
var gf = new GeneratingFunction(graph);
gf.ComputeGeneratingFunction();
var specEvalue = gf.GetSpectralEValue(scores.Score);
var rowStr = parser.GetRows()[i];
var items = rowStr.Split('\t').ToArray();
var newRowStr = string.Join("\t", items, 0, 15);
//writer.WriteLine("{0}\t{1}\t{2}", newRowStr, scores.Score, specEvalue);
lines[i] = string.Format("{0}\t{1}\t{2}", newRowStr, scores.Score, specEvalue);
//Console.WriteLine("{0}\t{1}\t{2}", items[0], scores.Score, specEvalue);
}
});
foreach (var line in (from item in lines where !string.IsNullOrWhiteSpace(item) select item).Take(20))
{
Console.WriteLine(line);
}
}
Console.WriteLine("Done");
}
}