private int FilterAndOutputFeatures(LcMsFeatureContainer container, LcMsPeakMatrix featureFinder, string outCsvFilePath, string ms1FeaturesFilePath)
{
var featureCounter = new int[1];
Ms1FtEntry.WriteToFile(ms1FeaturesFilePath, FilterFeaturesWithOutput(container, featureFinder, outCsvFilePath, featureCounter), Parameters.ScoreReport);
return(featureCounter[0]);
}
private int FilterAndOutputFeaturesOld(LcMsFeatureContainer container, LcMsPeakMatrix featureFinder, string outCsvFilePath, string ms1FeaturesFilePath)
{
var featureId = 0;
Stream csvStream = new MemoryStream();
if (Parameters.CsvOutput)
{
csvStream = new FileStream(outCsvFilePath, FileMode.Create, FileAccess.Write, FileShare.ReadWrite);
}
// write result files
using (var tsvWriter = new StreamWriter(ms1FeaturesFilePath))
using (var csvWriter = new StreamWriter(csvStream))
{
tsvWriter.WriteLine(GetHeaderString(Parameters.ScoreReport));
if (Parameters.CsvOutput)
{
csvWriter.WriteLine("scan_num,charge,abundance,mz,fit,monoisotopic_mw,FeatureID");
}
var filteredFeatures = container.GetFilteredFeatures(featureFinder);
foreach (var feature in filteredFeatures)
{
featureId++;
tsvWriter.WriteLine("{0}\t{1}", featureId, GetString(feature, Parameters.ScoreReport));
var mostAbuIdx = feature.TheoreticalEnvelope.IndexOrderByRanking[0];
if (Parameters.CsvOutput)
{
foreach (var envelope in feature.EnumerateEnvelopes())
{
//var mostAbuIsotopeInternalIndex = cluster.IsotopeList.SortedIndexByIntensity[0];
var mostAbuPeak = envelope.Peaks[mostAbuIdx];
if (mostAbuPeak == null || !mostAbuPeak.Active)
{
continue;
}
var fitscore = 1.0 - feature.BestCorrelationScore;
csvWriter.WriteLine("{0},{1},{2},{3},{4},{5},{6}", envelope.ScanNum, envelope.Charge, envelope.Abundance,
mostAbuPeak.Mz, fitscore, envelope.MonoMass, featureId);
}
}
}
}
return(featureId);
}
/// <summary>
/// Create <see cref="Ms1FtEntry"/> objects for the features, and output to csv (if desired).
/// </summary>
/// <param name="container"></param>
/// <param name="featureFinder"></param>
/// <param name="outCsvFilePath"></param>
/// <param name="featureCounter"></param>
/// <returns></returns>
private IEnumerable <Ms1FtEntry> FilterFeaturesWithOutput(LcMsFeatureContainer container, LcMsPeakMatrix featureFinder, string outCsvFilePath, int[] featureCounter)
{
// Using an array, since we can't use ref or out parameters
featureCounter[0] = 0;
Stream csvStream = new MemoryStream();
if (Parameters.CsvOutput)
{
csvStream = new FileStream(outCsvFilePath, FileMode.Create, FileAccess.Write, FileShare.ReadWrite);
}
using (var csvWriter = new StreamWriter(csvStream))
{
if (Parameters.CsvOutput)
{
csvWriter.WriteLine("scan_num,charge,abundance,mz,fit,monoisotopic_mw,FeatureID");
}
var filteredFeatures = container.GetFilteredFeatures(featureFinder);
foreach (var feature in filteredFeatures)
{
featureCounter[0]++;
if (Parameters.CsvOutput)
{
var mostAbuIdx = feature.TheoreticalEnvelope.IndexOrderByRanking[0];
foreach (var envelope in feature.EnumerateEnvelopes())
{
//var mostAbuIsotopeInternalIndex = cluster.IsotopeList.SortedIndexByIntensity[0];
var mostAbuPeak = envelope.Peaks[mostAbuIdx];
if (mostAbuPeak == null || !mostAbuPeak.Active)
{
continue;
}
var fitscore = 1.0 - feature.BestCorrelationScore;
csvWriter.WriteLine("{0},{1},{2},{3},{4},{5},{6}", envelope.ScanNum, envelope.Charge, envelope.Abundance,
mostAbuPeak.Mz, fitscore, envelope.MonoMass, featureCounter);
}
}
yield return(feature.ToMs1FtEntry(featureCounter[0]));
}
}
}
public void TestLcMsFeatureFinder()
{
var methodName = MethodBase.GetCurrentMethod().Name;
Utils.ShowStarting(methodName);
var pbfFilePath = Utils.GetPbfTestFilePath(false);
var pbfFile = Utils.GetTestFile(methodName, pbfFilePath);
// var outTsvFilePath = MassSpecDataReaderFactory.ChangeExtension(rawFile, "ms1ft");
//var scoreDataPath = @"D:\MassSpecFiles\training";
var scorer = new LcMsFeatureLikelihood();
var stopwatch = Stopwatch.StartNew();
Console.WriteLine(@"Start loading MS1 data from {0}", pbfFile.FullName);
var run = PbfLcMsRun.GetLcMsRun(pbfFile.FullName);
var featureFinder = new LcMsPeakMatrix(run, scorer);
Console.WriteLine(@"Complete loading MS1 data. Elapsed Time = {0:0.000} sec",
(stopwatch.ElapsedMilliseconds) / 1000.0d);
var container = new LcMsFeatureContainer(featureFinder.Ms1Spectra, scorer, new LcMsFeatureMergeComparer(new Tolerance(10)));
var minSearchMassBin = featureFinder.Comparer.GetBinNumber(11180.33677);
var maxSearchMassBin = featureFinder.Comparer.GetBinNumber(11180.33677);
double totalMassBin = maxSearchMassBin - minSearchMassBin + 1;
Console.WriteLine(@"Start MS1 feature extraction.");
stopwatch.Restart();
for (var binNum = minSearchMassBin; binNum <= maxSearchMassBin; binNum++)
{
var clusters = featureFinder.FindFeatures(binNum);
container.Add(clusters);
if (binNum > minSearchMassBin && (binNum - minSearchMassBin) % 1000 == 0)
{
var elapsed = (stopwatch.ElapsedMilliseconds) / 1000.0d;
var processedBins = binNum - minSearchMassBin;
var processedPercentage = ((double)processedBins / totalMassBin) * 100;
Console.WriteLine(
@"Processing {0:0.0}% of mass bins ({1:0.0} Da); elapsed time = {2:0.000} sec; # of features = {3}",
processedPercentage, featureFinder.Comparer.GetMzEnd(binNum), elapsed,
container.NumberOfFeatures);
}
}
Console.WriteLine(@"Complete MS1 feature extraction.");
Console.WriteLine(@" - Elapsed time = {0:0.000} sec", (stopwatch.ElapsedMilliseconds) / 1000.0d);
Console.WriteLine(@" - Number of extracted features = {0}", container.NumberOfFeatures);
// write result files
Console.WriteLine(@"Start selecting mutually independent features from feature network graph");
stopwatch.Stop();
// Start to quantify accurate abundance
stopwatch.Restart();
//var quantAnalyzer = new TargetMs1FeatureMatrix(run);
//var oriResult = new List<Ms1FeatureCluster>();
//var quantResult = new List<Ms1Feature>();
var featureId = 0;
var ms1ScanNums = run.GetMs1ScanVector();
//tsvWriter.WriteLine(GetHeaderString() + "\tQMinScanNum\tQMaxScanNum\tQMinCharge\tQMaxCharge\tQAbundance");
var filteredFeatures = container.GetFilteredFeatures(featureFinder);
foreach (var feature in filteredFeatures)
{
Console.Write(featureId);
Console.Write("\t");
Console.Write(feature.Mass);
Console.Write("\t");
Console.Write(feature.MinScanNum);
Console.Write("\t");
Console.Write(feature.MaxScanNum);
Console.Write("\t");
Console.Write(feature.MinCharge);
Console.Write("\t");
Console.Write(feature.MaxCharge);
Console.Write("\t");
Console.Write(feature.RepresentativeScanNum);
Console.Write("\t");
Console.Write(feature.RepresentativeMz);
Console.Write("\t");
Console.Write(feature.RepresentativeCharge);
Console.Write("\t");
//Console.Write(feature.BestSummedEnvelopeDistance); Console.Write("\t");
//Console.Write(feature.BestEnvelopeDistance); Console.Write("\t");
Console.Write(feature.BestDistanceScoreAcrossCharge[0]);
Console.Write("\t");
Console.Write(feature.BestDistanceScoreAcrossCharge[1]);
Console.Write("\t");
Console.Write(feature.BestCorrelationScoreAcrossCharge[0]);
Console.Write("\t");
Console.Write(feature.BestCorrelationScoreAcrossCharge[1]);
Console.Write("\t");
Console.Write(feature.BestIntensityScoreAcrossCharge[0]);
Console.Write("\t");
Console.Write(feature.BestIntensityScoreAcrossCharge[1]);
Console.Write("\t");
Console.Write(feature.AbundanceDistributionAcrossCharge[0]);
Console.Write("\t");
Console.Write(feature.AbundanceDistributionAcrossCharge[1]);
Console.Write("\t");
Console.Write(feature.XicCorrelationBetweenBestCharges[0]);
Console.Write("\t");
Console.Write(feature.XicCorrelationBetweenBestCharges[1]);
Console.Write("\t");
Console.Write(feature.Score);
Console.Write("\n");
featureId++;
}
}
/// <summary>
/// Find features in the data file
/// </summary>
/// <param name="rawFile">Data file (either a pbf file or a file type from which a pbf file can be auto-created)</param>
/// <returns>0 if success; negative number on error</returns>
private int ProcessFile(string rawFile)
{
var outDirectory = GetOutputDirectory(rawFile);
if (string.IsNullOrEmpty(outDirectory))
{
return(-1);
}
var baseName = Path.GetFileName(MassSpecDataReaderFactory.RemoveExtension(rawFile));
var ms1FeaturesFilePath = Path.Combine(outDirectory, baseName + "." + FileExtension);
var outCsvFilePath = Path.Combine(outDirectory, baseName + "_" + FileExtension + ".csv");
var pngFilePath = Path.Combine(outDirectory, baseName + "_" + FileExtension + ".png");
if (File.Exists(ms1FeaturesFilePath))
{
Console.WriteLine(@"ProMex output already exists: {0}", ms1FeaturesFilePath);
return(-2);
}
if (!File.Exists(rawFile))
{
ShowErrorMessage(@"Cannot find input file: " + rawFile);
return(-3);
}
var stopwatch = Stopwatch.StartNew();
Console.WriteLine(@"Start loading MS1 data from {0}", rawFile);
var run = PbfLcMsRun.GetLcMsRun(rawFile);
var featureFinder = new LcMsPeakMatrix(run, _likelihoodScorer, 1, 60, Parameters.MaxThreads);
Console.WriteLine(@"Complete loading MS1 data. Elapsed Time = {0:0.000} sec", (stopwatch.ElapsedMilliseconds) / 1000.0d);
if (run.GetMs1ScanVector().Length == 0)
{
ShowErrorMessage(@"Data file has no MS1 spectra: " + Path.GetFileName(rawFile));
return(-4);
}
var comparer = featureFinder.Comparer;
var container = new LcMsFeatureContainer(featureFinder.Ms1Spectra, _likelihoodScorer, new LcMsFeatureMergeComparer(new Tolerance(10)));
var minSearchMassBin = comparer.GetBinNumber(Parameters.MinSearchMass);
var maxSearchMassBin = comparer.GetBinNumber(Parameters.MaxSearchMass);
double totalMassBin = maxSearchMassBin - minSearchMassBin + 1;
Console.WriteLine(@"Start MS1 feature extraction.");
stopwatch.Restart();
for (var binNum = minSearchMassBin; binNum <= maxSearchMassBin; binNum++)
{
var clusters = featureFinder.FindFeatures(binNum);
container.Add(clusters);
if (binNum > minSearchMassBin && (binNum - minSearchMassBin) % 1000 == 0)
{
var elapsed = (stopwatch.ElapsedMilliseconds) / 1000.0d;
var processedBins = binNum - minSearchMassBin;
var processedPercentage = ((double)processedBins / totalMassBin) * 100;
Console.WriteLine(@"Processing {0:0.0}% of mass bins ({1:0.0} Da); elapsed time = {2:0.000} sec; # of features = {3}",
processedPercentage, featureFinder.Comparer.GetMzEnd(binNum), elapsed,
container.NumberOfFeatures);
}
}
Console.WriteLine(@"Complete MS1 feature extraction.");
Console.WriteLine(@" - Elapsed time = {0:0.000} sec", (stopwatch.ElapsedMilliseconds) / 1000.0d);
Console.WriteLine(@" - Number of extracted features = {0}", container.NumberOfFeatures);
Console.WriteLine(@"Start selecting mutually independent features from feature network graph");
stopwatch.Restart();
// write result files
var tsvWriter = new StreamWriter(ms1FeaturesFilePath);
tsvWriter.WriteLine(GetHeaderString(Parameters.ScoreReport));
StreamWriter csvWriter = null;
if (Parameters.CsvOutput)
{
csvWriter = new StreamWriter(outCsvFilePath);
csvWriter.WriteLine("scan_num,charge,abundance,mz,fit,monoisotopic_mw,FeatureID");
}
var filteredFeatures = container.GetFilteredFeatures(featureFinder);
var featureId = 0;
foreach (var feature in filteredFeatures)
{
featureId++;
tsvWriter.WriteLine("{0}\t{1}", featureId, GetString(feature, Parameters.ScoreReport));
var mostAbuIdx = feature.TheoreticalEnvelope.IndexOrderByRanking[0];
if (csvWriter != null)
{
foreach (var envelope in feature.EnumerateEnvelopes())
{
//var mostAbuIsotopeInternalIndex = cluster.IsotopeList.SortedIndexByIntensity[0];
var mostAbuPeak = envelope.Peaks[mostAbuIdx];
if (mostAbuPeak == null || !mostAbuPeak.Active)
{
continue;
}
var fitscore = 1.0 - feature.BestCorrelationScore;
csvWriter.WriteLine(string.Format("{0},{1},{2},{3},{4},{5},{6}", envelope.ScanNum, envelope.Charge, envelope.Abundance, mostAbuPeak.Mz, fitscore, envelope.MonoMass, featureId));
}
}
}
tsvWriter.Close();
Console.WriteLine(@"Complete feature filtration");
Console.WriteLine(@" - Elapsed time = {0:0.000} sec", (stopwatch.ElapsedMilliseconds) / 1000.0d);
Console.WriteLine(@" - Number of filtered features = {0}", featureId);
Console.WriteLine(@" - ProMex output: {0}", ms1FeaturesFilePath);
if (csvWriter != null)
{
csvWriter.Close();
Console.WriteLine(@" - ProMex output in ICR2LS format: {0}", outCsvFilePath);
}
if (Parameters.FeatureMapImage)
{
CreateFeatureMapImage(run, ms1FeaturesFilePath, pngFilePath);
}
return(0);
}
/// <summary>
/// Find features in the data file
/// </summary>
/// <param name="rawFile">Data file (either a pbf file or a file type from which a pbf file can be auto-created)</param>
/// <returns>0 if success; negative number on error</returns>
private int ProcessFile(string rawFile)
{
var outDirectory = GetOutputDirectory(rawFile);
if (string.IsNullOrEmpty(outDirectory))
return -1;
var baseName = Path.GetFileName(MassSpecDataReaderFactory.RemoveExtension(rawFile));
var ms1FeaturesFilePath = Path.Combine(outDirectory, baseName + "." + FileExtension);
var outCsvFilePath = Path.Combine(outDirectory, baseName + "_" + FileExtension + ".csv");
var pngFilePath = Path.Combine(outDirectory, baseName + "_" + FileExtension + ".png");
if (File.Exists(ms1FeaturesFilePath))
{
Console.WriteLine(@"ProMex output already exists: {0}", ms1FeaturesFilePath);
return -2;
}
if (!File.Exists(rawFile))
{
ShowErrorMessage(@"Cannot find input file: " + rawFile);
return -3;
}
var stopwatch = Stopwatch.StartNew();
Console.WriteLine(@"Start loading MS1 data from {0}", rawFile);
var run = PbfLcMsRun.GetLcMsRun(rawFile);
var featureFinder = new LcMsPeakMatrix(run, _likelihoodScorer, 1, 60, Parameters.MaxThreads);
Console.WriteLine(@"Complete loading MS1 data. Elapsed Time = {0:0.000} sec", (stopwatch.ElapsedMilliseconds) / 1000.0d);
if (run.GetMs1ScanVector().Length == 0)
{
ShowErrorMessage(@"Data file has no MS1 spectra: " + Path.GetFileName(rawFile));
return -4;
}
var comparer = featureFinder.Comparer;
var container = new LcMsFeatureContainer(featureFinder.Ms1Spectra, _likelihoodScorer, new LcMsFeatureMergeComparer(new Tolerance(10)));
var minSearchMassBin = comparer.GetBinNumber(Parameters.MinSearchMass);
var maxSearchMassBin = comparer.GetBinNumber(Parameters.MaxSearchMass);
double totalMassBin = maxSearchMassBin - minSearchMassBin + 1;
Console.WriteLine(@"Start MS1 feature extraction.");
stopwatch.Restart();
for (var binNum = minSearchMassBin; binNum <= maxSearchMassBin; binNum++)
{
var clusters = featureFinder.FindFeatures(binNum);
container.Add(clusters);
if (binNum > minSearchMassBin && (binNum - minSearchMassBin) % 1000 == 0)
{
var elapsed = (stopwatch.ElapsedMilliseconds) / 1000.0d;
var processedBins = binNum - minSearchMassBin;
var processedPercentage = ((double)processedBins / totalMassBin) * 100;
Console.WriteLine(@"Processing {0:0.0}% of mass bins ({1:0.0} Da); elapsed time = {2:0.000} sec; # of features = {3}",
processedPercentage, featureFinder.Comparer.GetMzEnd(binNum), elapsed,
container.NumberOfFeatures);
}
}
Console.WriteLine(@"Complete MS1 feature extraction.");
Console.WriteLine(@" - Elapsed time = {0:0.000} sec", (stopwatch.ElapsedMilliseconds) / 1000.0d);
Console.WriteLine(@" - Number of extracted features = {0}", container.NumberOfFeatures);
Console.WriteLine(@"Start selecting mutually independent features from feature network graph");
stopwatch.Restart();
// write result files
var tsvWriter = new StreamWriter(ms1FeaturesFilePath);
tsvWriter.WriteLine(GetHeaderString(Parameters.ScoreReport));
StreamWriter csvWriter = null;
if (Parameters.CsvOutput)
{
csvWriter = new StreamWriter(outCsvFilePath);
csvWriter.WriteLine("scan_num,charge,abundance,mz,fit,monoisotopic_mw,FeatureID");
}
var filteredFeatures = container.GetFilteredFeatures(featureFinder);
var featureId = 0;
foreach (var feature in filteredFeatures)
{
featureId++;
tsvWriter.WriteLine("{0}\t{1}", featureId, GetString(feature, Parameters.ScoreReport));
var mostAbuIdx = feature.TheoreticalEnvelope.IndexOrderByRanking[0];
if (csvWriter != null)
{
foreach (var envelope in feature.EnumerateEnvelopes())
{
//var mostAbuIsotopeInternalIndex = cluster.IsotopeList.SortedIndexByIntensity[0];
var mostAbuPeak = envelope.Peaks[mostAbuIdx];
if (mostAbuPeak == null || !mostAbuPeak.Active) continue;
var fitscore = 1.0 - feature.BestCorrelationScore;
csvWriter.WriteLine(string.Format("{0},{1},{2},{3},{4},{5},{6}", envelope.ScanNum, envelope.Charge, envelope.Abundance, mostAbuPeak.Mz, fitscore, envelope.MonoMass, featureId));
}
}
}
tsvWriter.Close();
Console.WriteLine(@"Complete feature filtration");
Console.WriteLine(@" - Elapsed time = {0:0.000} sec", (stopwatch.ElapsedMilliseconds) / 1000.0d);
Console.WriteLine(@" - Number of filtered features = {0}", featureId);
Console.WriteLine(@" - ProMex output: {0}", ms1FeaturesFilePath);
if (csvWriter != null)
{
csvWriter.Close();
Console.WriteLine(@" - ProMex output in ICR2LS format: {0}", outCsvFilePath);
}
if (Parameters.FeatureMapImage)
{
CreateFeatureMapImage(run, ms1FeaturesFilePath, pngFilePath);
}
return 0;
}
public void TestLcMsFeatureFinder()
{
var methodName = MethodBase.GetCurrentMethod().Name;
TestUtils.ShowStarting(methodName);
const string rawFile = @"D:\MassSpecFiles\training\raw\QC_Shew_Intact_26Sep14_Bane_C2Column3.pbf";
//const string rawFile = @"D:\MassSpecFiles\CompRef\CPTAC_Intact_CR_Pool_2_25Jun15_Bane_15-02-02RZ.pbf";
//const string rawFile = @"D:\MassSpecFiles\IMER\Dey_IMERblast_01_08May14_Alder_14-01-33.pbf";
//const string rawFile = @"\\proto-11\MSXML_Cache\PBF_Gen_1_193\2015_3\MZ20150729FG_WT1.pbf";
if (!File.Exists(rawFile))
{
Assert.Ignore(@"Skipping test {0} since file not found: {1}", methodName, rawFile);
}
// var outTsvFilePath = MassSpecDataReaderFactory.ChangeExtension(rawFile, "ms1ft");
//var scoreDataPath = @"D:\MassSpecFiles\training";
var scorer = new LcMsFeatureLikelihood();
var stopwatch = Stopwatch.StartNew();
Console.WriteLine(@"Start loading MS1 data from {0}", rawFile);
var run = PbfLcMsRun.GetLcMsRun(rawFile);
var featureFinder = new LcMsPeakMatrix(run, scorer);
Console.WriteLine(@"Complete loading MS1 data. Elapsed Time = {0:0.000} sec",
(stopwatch.ElapsedMilliseconds)/1000.0d);
var container = new LcMsFeatureContainer(featureFinder.Ms1Spectra, scorer, new LcMsFeatureMergeComparer(new Tolerance(10)));
var minSearchMassBin = featureFinder.Comparer.GetBinNumber(11180.33677);
var maxSearchMassBin = featureFinder.Comparer.GetBinNumber(11180.33677);
double totalMassBin = maxSearchMassBin - minSearchMassBin + 1;
Console.WriteLine(@"Start MS1 feature extraction.");
stopwatch.Restart();
for (var binNum = minSearchMassBin; binNum <= maxSearchMassBin; binNum++)
{
var clusters = featureFinder.FindFeatures(binNum);
container.Add(clusters);
if (binNum > minSearchMassBin && (binNum - minSearchMassBin)%1000 == 0)
{
var elapsed = (stopwatch.ElapsedMilliseconds)/1000.0d;
var processedBins = binNum - minSearchMassBin;
var processedPercentage = ((double) processedBins/totalMassBin)*100;
Console.WriteLine(
@"Processing {0:0.0}% of mass bins ({1:0.0} Da); elapsed time = {2:0.000} sec; # of features = {3}",
processedPercentage, featureFinder.Comparer.GetMzEnd(binNum), elapsed,
container.NumberOfFeatures);
}
}
Console.WriteLine(@"Complete MS1 feature extraction.");
Console.WriteLine(@" - Elapsed time = {0:0.000} sec", (stopwatch.ElapsedMilliseconds)/1000.0d);
Console.WriteLine(@" - Number of extracted features = {0}", container.NumberOfFeatures);
// write result files
Console.WriteLine(@"Start selecting mutually independent features from feature network graph");
stopwatch.Stop();
// Start to quantify accurate abundance
stopwatch.Restart();
//var quantAnalyzer = new TargetMs1FeatureMatrix(run);
//var oriResult = new List<Ms1FeatureCluster>();
//var quantResult = new List<Ms1Feature>();
var featureId = 0;
var ms1ScanNums = run.GetMs1ScanVector();
//tsvWriter.WriteLine(GetHeaderString() + "\tQMinScanNum\tQMaxScanNum\tQMinCharge\tQMaxCharge\tQAbundance");
var filteredFeatures = container.GetFilteredFeatures(featureFinder);
foreach (var feature in filteredFeatures)
{
Console.Write(featureId);
Console.Write("\t");
Console.Write(feature.Mass);
Console.Write("\t");
Console.Write(feature.MinScanNum);
Console.Write("\t");
Console.Write(feature.MaxScanNum);
Console.Write("\t");
Console.Write(feature.MinCharge);
Console.Write("\t");
Console.Write(feature.MaxCharge);
Console.Write("\t");
Console.Write(feature.RepresentativeScanNum);
Console.Write("\t");
Console.Write(feature.RepresentativeMz);
Console.Write("\t");
Console.Write(feature.RepresentativeCharge);
Console.Write("\t");
//Console.Write(feature.BestSummedEnvelopeDistance); Console.Write("\t");
//Console.Write(feature.BestEnvelopeDistance); Console.Write("\t");
Console.Write(feature.BestDistanceScoreAcrossCharge[0]);
Console.Write("\t");
Console.Write(feature.BestDistanceScoreAcrossCharge[1]);
Console.Write("\t");
Console.Write(feature.BestCorrelationScoreAcrossCharge[0]);
Console.Write("\t");
Console.Write(feature.BestCorrelationScoreAcrossCharge[1]);
Console.Write("\t");
Console.Write(feature.BestIntensityScoreAcrossCharge[0]);
Console.Write("\t");
Console.Write(feature.BestIntensityScoreAcrossCharge[1]);
Console.Write("\t");
Console.Write(feature.AbundanceDistributionAcrossCharge[0]);
Console.Write("\t");
Console.Write(feature.AbundanceDistributionAcrossCharge[1]);
Console.Write("\t");
Console.Write(feature.XicCorrelationBetweenBestCharges[0]);
Console.Write("\t");
Console.Write(feature.XicCorrelationBetweenBestCharges[1]);
Console.Write("\t");
Console.Write(feature.Score);
Console.Write("\n");
featureId++;
}
}
public void TestLcMsFeatureFinder()
{
var methodName = MethodBase.GetCurrentMethod().Name;
TestUtils.ShowStarting(methodName);
const string rawFile = @"D:\MassSpecFiles\training\raw\QC_Shew_Intact_26Sep14_Bane_C2Column3.pbf";
//const string rawFile = @"D:\MassSpecFiles\CompRef\CPTAC_Intact_CR_Pool_2_25Jun15_Bane_15-02-02RZ.pbf";
//const string rawFile = @"D:\MassSpecFiles\IMER\Dey_IMERblast_01_08May14_Alder_14-01-33.pbf";
//const string rawFile = @"\\proto-11\MSXML_Cache\PBF_Gen_1_193\2015_3\MZ20150729FG_WT1.pbf";
if (!File.Exists(rawFile))
{
Assert.Ignore(@"Skipping test {0} since file not found: {1}", methodName, rawFile);
}
// var outTsvFilePath = MassSpecDataReaderFactory.ChangeExtension(rawFile, "ms1ft");
//var scoreDataPath = @"D:\MassSpecFiles\training";
var scorer = new LcMsFeatureLikelihood();
var stopwatch = Stopwatch.StartNew();
Console.WriteLine(@"Start loading MS1 data from {0}", rawFile);
var run = PbfLcMsRun.GetLcMsRun(rawFile);
var featureFinder = new LcMsPeakMatrix(run, scorer);
Console.WriteLine(@"Complete loading MS1 data. Elapsed Time = {0:0.000} sec",
(stopwatch.ElapsedMilliseconds) / 1000.0d);
var container = new LcMsFeatureContainer(featureFinder.Ms1Spectra, scorer, new LcMsFeatureMergeComparer(new Tolerance(10)));
var minSearchMassBin = featureFinder.Comparer.GetBinNumber(11180.33677);
var maxSearchMassBin = featureFinder.Comparer.GetBinNumber(11180.33677);
double totalMassBin = maxSearchMassBin - minSearchMassBin + 1;
Console.WriteLine(@"Start MS1 feature extraction.");
stopwatch.Restart();
for (var binNum = minSearchMassBin; binNum <= maxSearchMassBin; binNum++)
{
var clusters = featureFinder.FindFeatures(binNum);
container.Add(clusters);
if (binNum > minSearchMassBin && (binNum - minSearchMassBin) % 1000 == 0)
{
var elapsed = (stopwatch.ElapsedMilliseconds) / 1000.0d;
var processedBins = binNum - minSearchMassBin;
var processedPercentage = ((double)processedBins / totalMassBin) * 100;
Console.WriteLine(
@"Processing {0:0.0}% of mass bins ({1:0.0} Da); elapsed time = {2:0.000} sec; # of features = {3}",
processedPercentage, featureFinder.Comparer.GetMzEnd(binNum), elapsed,
container.NumberOfFeatures);
}
}
Console.WriteLine(@"Complete MS1 feature extraction.");
Console.WriteLine(@" - Elapsed time = {0:0.000} sec", (stopwatch.ElapsedMilliseconds) / 1000.0d);
Console.WriteLine(@" - Number of extracted features = {0}", container.NumberOfFeatures);
// write result files
Console.WriteLine(@"Start selecting mutually independent features from feature network graph");
stopwatch.Stop();
// Start to quantify accurate abundance
stopwatch.Restart();
//var quantAnalyzer = new TargetMs1FeatureMatrix(run);
//var oriResult = new List<Ms1FeatureCluster>();
//var quantResult = new List<Ms1Feature>();
var featureId = 0;
var ms1ScanNums = run.GetMs1ScanVector();
//tsvWriter.WriteLine(GetHeaderString() + "\tQMinScanNum\tQMaxScanNum\tQMinCharge\tQMaxCharge\tQAbundance");
var filteredFeatures = container.GetFilteredFeatures(featureFinder);
foreach (var feature in filteredFeatures)
{
Console.Write(featureId);
Console.Write("\t");
Console.Write(feature.Mass);
Console.Write("\t");
Console.Write(feature.MinScanNum);
Console.Write("\t");
Console.Write(feature.MaxScanNum);
Console.Write("\t");
Console.Write(feature.MinCharge);
Console.Write("\t");
Console.Write(feature.MaxCharge);
Console.Write("\t");
Console.Write(feature.RepresentativeScanNum);
Console.Write("\t");
Console.Write(feature.RepresentativeMz);
Console.Write("\t");
Console.Write(feature.RepresentativeCharge);
Console.Write("\t");
//Console.Write(feature.BestSummedEnvelopeDistance); Console.Write("\t");
//Console.Write(feature.BestEnvelopeDistance); Console.Write("\t");
Console.Write(feature.BestDistanceScoreAcrossCharge[0]);
Console.Write("\t");
Console.Write(feature.BestDistanceScoreAcrossCharge[1]);
Console.Write("\t");
Console.Write(feature.BestCorrelationScoreAcrossCharge[0]);
Console.Write("\t");
Console.Write(feature.BestCorrelationScoreAcrossCharge[1]);
Console.Write("\t");
Console.Write(feature.BestIntensityScoreAcrossCharge[0]);
Console.Write("\t");
Console.Write(feature.BestIntensityScoreAcrossCharge[1]);
Console.Write("\t");
Console.Write(feature.AbundanceDistributionAcrossCharge[0]);
Console.Write("\t");
Console.Write(feature.AbundanceDistributionAcrossCharge[1]);
Console.Write("\t");
Console.Write(feature.XicCorrelationBetweenBestCharges[0]);
Console.Write("\t");
Console.Write(feature.XicCorrelationBetweenBestCharges[1]);
Console.Write("\t");
Console.Write(feature.Score);
Console.Write("\n");
featureId++;
}
}
/// <summary>
/// Find features in the data file
/// </summary>
/// <param name="rawFile">Data file (either a pbf file or a file type from which a pbf file can be auto-created)</param>
/// <returns>0 if success; negative number on error</returns>
private int ProcessFile(string rawFile)
{
var outDirectory = GetOutputDirectory(rawFile);
if (string.IsNullOrEmpty(outDirectory))
{
return(-1);
}
var baseName = Path.GetFileName(MassSpecDataReaderFactory.RemoveExtension(rawFile));
var ms1FeaturesFilePath = Path.Combine(outDirectory, baseName + "." + FileExtension);
var outCsvFilePath = Path.Combine(outDirectory, baseName + "_" + FileExtension + ".csv");
var pngFilePath = Path.Combine(outDirectory, baseName + "_" + FileExtension + ".png");
if (File.Exists(ms1FeaturesFilePath))
{
ShowErrorMessage("ProMex output already exists: " + ms1FeaturesFilePath);
return(-2);
}
if (!File.Exists(rawFile))
{
ShowErrorMessage("Cannot find input file: " + rawFile);
return(-3);
}
var stopwatch = Stopwatch.StartNew();
Console.WriteLine("Start loading MS1 data from {0}", rawFile);
var run = PbfLcMsRun.GetLcMsRun(rawFile);
var featureFinder = new LcMsPeakMatrix(run, _likelihoodScorer, 1, 60, Parameters.MaxThreads);
Console.WriteLine("Complete loading MS1 data. Elapsed Time = {0:0.000} sec", (stopwatch.ElapsedMilliseconds) / 1000.0d);
if (run.GetMs1ScanVector().Length == 0)
{
ShowErrorMessage(@"Data file has no MS1 spectra: " + Path.GetFileName(rawFile));
return(-4);
}
if (featureFinder.Ms1PeakCount == 0)
{
ShowErrorMessage(@"Data file has no MS1 peaks: " + Path.GetFileName(rawFile));
return(-5);
}
var comparer = featureFinder.Comparer;
var container = new LcMsFeatureContainer(featureFinder.Ms1Spectra, _likelihoodScorer, new LcMsFeatureMergeComparer(new Tolerance(10)));
var minSearchMassBin = comparer.GetBinNumber(Parameters.MinSearchMass);
var maxSearchMassBin = comparer.GetBinNumber(Parameters.MaxSearchMass);
double totalMassBin = maxSearchMassBin - minSearchMassBin + 1;
Console.WriteLine("Start MS1 feature extraction.");
stopwatch.Restart();
for (var binNum = minSearchMassBin; binNum <= maxSearchMassBin; binNum++)
{
var clusters = featureFinder.FindFeatures(binNum);
container.Add(clusters);
if (binNum > minSearchMassBin && (binNum - minSearchMassBin) % 1000 == 0)
{
var elapsed = (stopwatch.ElapsedMilliseconds) / 1000.0d;
var processedBins = binNum - minSearchMassBin;
var processedPercentage = processedBins / totalMassBin * 100;
Console.WriteLine("Processing {0:0.0}% of mass bins ({1:0.0} Da); elapsed time = {2:0.000} sec; # of features = {3}",
processedPercentage, featureFinder.Comparer.GetMzEnd(binNum), elapsed,
container.NumberOfFeatures);
}
}
Console.WriteLine("Complete MS1 feature extraction.");
Console.WriteLine(" - Elapsed time = {0:0.000} sec", (stopwatch.ElapsedMilliseconds) / 1000.0d);
Console.WriteLine(" - Number of extracted features = {0}", container.NumberOfFeatures);
Console.WriteLine("Start selecting mutually independent features from feature network graph");
stopwatch.Restart();
var featureId = FilterAndOutputFeatures(container, featureFinder, outCsvFilePath, ms1FeaturesFilePath);
Console.WriteLine("Complete feature filtration");
Console.WriteLine(" - Elapsed time = {0:0.000} sec", (stopwatch.ElapsedMilliseconds) / 1000.0d);
Console.WriteLine(" - Number of filtered features = {0}", featureId);
Console.WriteLine(" - ProMex output: {0}", ms1FeaturesFilePath);
if (Parameters.CsvOutput)
{
Console.WriteLine(" - ProMex output in ICR2LS format: {0}", outCsvFilePath);
}
if (Parameters.FeatureMapImage)
{
CreateFeatureMapImage(run, ms1FeaturesFilePath, pngFilePath);
}
return(0);
}
