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 AddMostAbundantIsotopePeakIntensity()
{
var methodName = MethodBase.GetCurrentMethod().Name;
TestUtils.ShowStarting(methodName);
const string rawFilePath = @"\\proto-2\UnitTest_Files\InformedProteomics_TestFiles\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);
const string resultFilePath = @"\\proto-2\UnitTest_Files\InformedProteomics_TestFiles\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
const string newResultFilePath = @"\\proto-2\UnitTest_Files\InformedProteomics_TestFiles\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");
}
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);
}
}
}
public string ProcessFile(string rawFile, string resultFile, string methodName)
{
if (!File.Exists(rawFile))
{
Console.WriteLine(@"Warning: Skipping test {0} since file not found: {1}", methodName, rawFile);
return "\n";
}
if (!File.Exists(resultFile))
{
Console.WriteLine(@"Warning: Skipping test {0} since file not found: {1}", methodName, resultFile);
return "\n";
}
var tsvParser = new TsvFileParser(resultFile);
var headerList = tsvParser.GetHeaders();
var tsvData = tsvParser.GetAllData();
var ms2ScanNumbers = tsvData["Scan"];
var run = PbfLcMsRun.GetLcMsRun(rawFile, 0, 0);
var resultLine = "";
for (int i = 0; i < ms2ScanNumbers.Count; i++)
{
var scanNum = Int32.Parse(ms2ScanNumbers[i]);
var spectrum = run.GetSpectrum(scanNum) as ProductSpectrum;
int tsvIndex = ms2ScanNumbers.FindIndex(x => Int32.Parse(x) == scanNum);
var qValue = Double.Parse(tsvData["QValue"].ElementAt(tsvIndex));
if (qValue > 0.01) continue;
var seqStr = tsvData["Sequence"].ElementAt(tsvIndex).Trim();
var seqMod = tsvData["Modifications"].ElementAt(tsvIndex).Trim();
var matchedFrags = tsvData["#MatchedFragments"].ElementAt(tsvIndex).Trim();
var aaSet = new AminoAcidSet();
var sequence = Sequence.CreateSequence(seqStr, seqMod, aaSet);
var tol = new Tolerance(10);
var sequenceFinder = new SequenceTagIndexFinder(tol, 1, 10);
var results = sequenceFinder.GetLongestSequence(spectrum, sequence);
resultLine += String.Format("{0},{1},{2},{3},{4},{5},{6},{7},{8},\n", scanNum, matchedFrags, seqStr, results.Item1, results.Item2,results.Item3,results.Item4,results.Item5,results.Item6);
}
return resultLine;
}
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 TestSequenceTag()
{
var methodName = MethodBase.GetCurrentMethod().Name;
TestUtils.ShowStarting(methodName);
//const string TestRawFile = @"D:\\Vlad_TopDown\\raw\\yufeng_column_test2.raw";
//const string TestResultFile = @"D:\\Vlad_TopDown\\results\\yufeng_column_test2_IcTda.tsv";
const string TestRawFile = @"D:\MassSpecFiles\training\raw\QC_Shew_Intact_26Sep14_Bane_C2Column3.pbf";
const string TestResultFile = @"D:\MassSpecFiles\training\IdResult\QC_Shew_Intact_26Sep14_Bane_C2Column3_IcTda.tsv";
//const string TestRawFile = @"D:\MassSpecFiles\Lewy\Lewy_intact_01.raw";
//const string TestResultFile = @"D:\MassSpecFiles\Lewy\Lewy_intact_01_IcTda.tsv";
if (!File.Exists(TestRawFile))
{
Assert.Ignore(@"Skipping test {0} since file not found: {1}", methodName, TestRawFile);
}
if (!File.Exists(TestResultFile))
{
Assert.Ignore(@"Skipping test {0} since file not found: {1}", methodName, TestResultFile);
}
// Configure amino acid set
var aminoAcidList = new List<AminoAcid>();
foreach (var aa in AminoAcid.StandardAminoAcidArr)
{
aminoAcidList.Add(aa);
aminoAcidList.Add(new ModifiedAminoAcid(aa, Modification.Acetylation));
aminoAcidList.Add(new ModifiedAminoAcid(aa, Modification.Oxidation));
}
//const int MaxTags = 100000;
var tsvParser = new TsvFileParser(TestResultFile);
var headerList = tsvParser.GetHeaders();
var tsvData = tsvParser.GetAllData();
var ms2ScanNumbers = tsvData["Scan"];
var run = PbfLcMsRun.GetLcMsRun(TestRawFile);
var nSpec = 0;
var nHitSpec = 0;
for (var i = 0; i < ms2ScanNumbers.Count; i++)
//foreach(var scanNum in targetScans)
{
var scanNum = Int32.Parse(ms2ScanNumbers[i]);
//if (scanNum != 4672) continue;
var spectrum = run.GetSpectrum(scanNum) as ProductSpectrum;
int tsvIndex = ms2ScanNumbers.FindIndex(x => Int32.Parse(x) == scanNum);
var qValue = double.Parse(tsvData["QValue"].ElementAt(tsvIndex));
if (qValue > 0.01) break;
var seqStr = tsvData["Sequence"].ElementAt(tsvIndex).Trim();
var modStr = tsvData["Modifications"].ElementAt(tsvIndex).Trim();
var tolerance = new Tolerance(5);
var tagFinder = new SequenceTagFinder(spectrum, tolerance, 5, 8, aminoAcidList.ToArray());
var nTags = 0;
var nHit = 0;
var seqOjb = Sequence.CreateSequence(seqStr, modStr, new AminoAcidSet());
var compWithoutH2O = seqOjb.Composition - Composition.H2O;
//Console.WriteLine(compWithoutH2O.Mass);
foreach (var seqTagStr in tagFinder.GetAllSequenceTagString())
{
if (seqStr.Contains(seqTagStr.Sequence)) //|| seqStr.Contains(Reverse(tagStr)))
{
//var idx = seqStr.IndexOf(seqTagStr.Sequence);
//seqStr.Substring(0, idx)
/*var comp2 = seqOjb.GetComposition(0, idx);
Console.Write(comp2.Mass);
Console.Write("\t");
Console.Write(seqTagStr.FlankingMass);
Console.Write("\t");
Console.Write(seqTagStr.Sequence);
Console.Write("\t");
Console.Write(seqTagStr.IsPrefix);
Console.WriteLine("");
*/
if (seqStr.Contains(seqTagStr.Sequence)) nHit++;
}
nTags++;
}
nSpec++;
if (nHit > 0) nHitSpec++;
Console.WriteLine(@"[{0}]seqLen = {1}: {2}/{3}", scanNum, seqStr.Length, nHit, nTags);
}
//var existingTags = tagFinder.ExtractExistingSequneceTags(sequence);
Console.Write("{0}/{1}", nHitSpec, nSpec);
}
public void CreatePeptideAbundanceTableWithSkyline()
{
var methodName = MethodBase.GetCurrentMethod().Name;
TestUtils.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 ProcessIprg2015PreStudy()
{
var methodName = MethodBase.GetCurrentMethod().Name;
TestUtils.ShowStarting(methodName);
const string dir = @"H:\Research\IPRG2015";
const string databaseFilePath = dir + @"\database\yeast6proteaprotein.fasta";
if (!File.Exists(databaseFilePath))
{
Assert.Ignore(@"Skipping test {0} since file not found: {1}", methodName, databaseFilePath);
}
var database = new FastaDatabase(databaseFilePath);
database.Read();
const string jobFilePath = dir + @"\Jobs.tsv";
if (!File.Exists(jobFilePath))
{
Assert.Ignore(@"Skipping test {0} since file not found: {1}", methodName, jobFilePath);
}
var jobParser = new TsvFileParser(jobFilePath);
var jobs = jobParser.GetData("Jobs").Select(j => Convert.ToInt32(j)).ToArray();
var experiments = jobParser.GetData("Experiments").Select(e => e.Split('_')[2]).ToArray();
//const string resultFilePath = dir + @"\AMT_Proteins_NA.tsv";
//const string outputFilePath = dir + @"\AMT_Proteins.tsv";
const string resultFilePath = dir + @"\AMT_Peptides_NA.tsv";
const string outputFilePath = dir + @"\AMT_Peptides.tsv";
var parser = new TsvFileParser(resultFilePath);
var headers = parser.GetHeaders();
var jobColNum = new int[jobs.Length];
for (var i = 0; i < jobs.Length; i++)
{
for (var j = 0; j < headers.Count; j++)
{
if (headers[j].Contains("" + jobs[i]))
{
jobColNum[i] = j;
break;
}
}
}
for (var i = 0; i < jobs.Length; i++)
{
Console.WriteLine("{0}\t{1}\t{2}", jobs[i], jobColNum[i], experiments[i]);
}
using (var writer = new StreamWriter(outputFilePath))
{
var peptides = parser.GetData("Peptide"); // Peptides
var proteins = parser.GetData("Reference"); // Proteins
var abundances = new string[jobs.Length][];
for (var i = 0; i < jobs.Length; i++)
{
abundances[i] = parser.GetData(headers[jobColNum[i]]).ToArray();
}
if (peptides != null) writer.Write("Peptide\t");
writer.Write("Protein\tLength");
for (var i = 0; i < jobs.Length; i++)
{
writer.Write("\t" + experiments[i]);
}
writer.WriteLine("\tSpikeIn");
for (var i = 0; i < proteins.Count; i++)
{
var protein = proteins[i];
if (protein.StartsWith("XXX") || protein.StartsWith("Contaminant")) continue;
var length = database.GetProteinLength(protein);
//if (length <= 0)
//{
// Console.WriteLine("Shit!");
// return;
//}
if (peptides != null) writer.Write(peptides[i] + "\t");
writer.Write(protein + "\t" + length);
for (var j = 0; j < jobs.Length; j++)
{
writer.Write("\t" + abundances[j][i]);
}
writer.WriteLine("\t" + (protein.StartsWith("STANDARD") ? 1 : 0));
}
}
}