public LcMsFeatureContainer(List <Ms1Spectrum> ms1Spectra, LcMsFeatureLikelihood scorer, INodeComparer <LcMsPeakCluster> mergeComparer)
{
_featureList = new List <LcMsPeakCluster>();
_spectra = ms1Spectra;
_scorer = scorer;
_mergeComparer = mergeComparer;
}
/// <summary>
/// Constructor
/// </summary>
/// <param name="parameters"></param>
public LcMsFeatureFinderLauncher(LcMsFeatureFinderInputParameters parameters = null)
{
Parameters = parameters ?? new LcMsFeatureFinderInputParameters();
try
{
_likelihoodScorer = new LcMsFeatureLikelihood(Parameters.LikelihoodScoreThreshold);
}
catch (FileNotFoundException fe)
{
ShowErrorMessage(fe.Message);
}
}
public void TestFeatureExampleForFigure()
{
var methodName = MethodBase.GetCurrentMethod().Name;
Utils.ShowStarting(methodName);
const string rawFile = @"\\proto-11\MSXML_Cache\PBF_Gen_1_193\2015_1\CPTAC_Intact_rep6_15Jan15_Bane_C2-14-08-02RZ.pbf";
//const string rawFile = @"D:\MassSpecFiles\training\raw\QC_Shew_Intact_26Sep14_Bane_C2Column3.pbf";
if (!File.Exists(rawFile))
{
Assert.Ignore(@"Skipping test {0} since file not found: {1}", methodName, rawFile);
}
var run = PbfLcMsRun.GetLcMsRun(rawFile);
var scorer = new LcMsFeatureLikelihood();
var featureFinder = new LcMsPeakMatrix(run, scorer);
var feature = featureFinder.GetLcMsPeakCluster(28061.6177, 20, 34, 7624, 7736);
var resultsFilePath = Path.Combine(Path.GetTempPath(), Path.GetFileNameWithoutExtension(rawFile) + "_peaks.txt");
var writer = new StreamWriter(resultsFilePath);
writer.Write("{0}\t{1}\t{2}\t{3}\t{4}\t{5}\t{6}\n", "Scan", "Elution_Time", "Charge", "ID", "MZ", "Intensity", "Pearson_Correlation");
var envelope = feature.TheoreticalEnvelope;
foreach (var e in envelope.Isotopes)
{
Console.WriteLine(e.Ratio);
}
foreach (var env in feature.EnumerateEnvelopes())
{
var corr = env.PearsonCorrelation;
for (var i = 0; i < envelope.Size; i++)
{
var peak = env.Peaks[i];
if (peak == null)
{
continue;
}
writer.Write("{0}\t{1}\t{2}\t{3}\t{4}\t{5}\t{6}\n", env.ScanNum, run.GetElutionTime(env.ScanNum), env.Charge, i, peak.Mz, peak.Intensity, corr);
}
}
writer.Close();
Console.WriteLine("Results are in file " + resultsFilePath);
}
public void TestLcMsFeatureXic()
{
var methodName = MethodBase.GetCurrentMethod().Name;
Utils.ShowStarting(methodName);
const string rawFile = @"\\proto-11\MSXML_Cache\PBF_Gen_1_193\2015_1\CPTAC_Intact_rep2_15Jan15_Bane_C2-14-08-02RZ.pbf";
//const string rawFile = @"D:\MassSpecFiles\training\raw\QC_Shew_Intact_26Sep14_Bane_C2Column3.pbf";
if (!File.Exists(rawFile))
{
Assert.Ignore(@"Skipping test {0} since file not found: {1}", methodName, rawFile);
}
var run = PbfLcMsRun.GetLcMsRun(rawFile);
var scorer = new LcMsFeatureLikelihood();
var featureFinder = new LcMsPeakMatrix(run, scorer);
var feature = featureFinder.GetLcMsPeakCluster(2388.278, 4, 3774, 3907);
//feature = featureFinder.GetLcMsPeakCluster(8151.3706, 7, 13, 4201, 4266);
//feature = featureFinder.GetLcMsPeakCluster(8151.41789, 7, 13, 2861, 2941);
var ms1ScanToIndex = run.GetMs1ScanNumToIndex();
var minCol = ms1ScanToIndex[feature.MinScanNum];
var maxCol = ms1ScanToIndex[feature.MaxScanNum];
//var minRow = feature.MinCharge - LcMsPeakMatrix.MinScanCharge;
//var maxRow = feature.MaxCharge - LcMsPeakMatrix.MinScanCharge;
Console.WriteLine("---------------------------------------------------------------");
for (var i = 0; i < feature.Envelopes.Length; i++)
{
for (var j = 0; j < feature.Envelopes[i].Length; j++)
{
Console.Write(feature.Envelopes[i][j] != null ? feature.Envelopes[i][j].PearsonCorrelation : 0);
Console.Write("\t");
}
Console.Write("\n");
}
Console.WriteLine("---------------------------------------------------------------");
for (var i = 0; i < feature.Envelopes.Length; i++)
{
for (var j = 0; j < feature.Envelopes[i].Length; j++)
{
Console.Write(feature.Envelopes[i][j] != null ? feature.Envelopes[i][j].BhattacharyyaDistance : 0);
Console.Write("\t");
}
Console.Write("\n");
}
Console.WriteLine("---------------------------------------------------------------");
for (var i = 0; i < feature.Envelopes.Length; i++)
{
for (var j = 0; j < feature.Envelopes[i].Length; j++)
{
Console.Write(feature.Envelopes[i][j] != null ? feature.Envelopes[i][j].Abundance : 0);
Console.Write("\t");
}
Console.Write("\n");
}
}
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++;
}
}
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++;
}
}
