public LipidGroupSearchResult(LipidTarget lipidTarget, List <Lipid> lipidList, SpectrumSearchResult spectrumSearchResult, ScoreModel scoreModel)
{
LipidTarget = lipidTarget;
LipidList = lipidList;
SpectrumSearchResult = spectrumSearchResult;
ShouldExport = false;
Score = scoreModel.ScoreLipid(this);
if (spectrumSearchResult.HcdSpectrum != null)
{
DisplayScanNum = spectrumSearchResult.HcdSpectrum.ScanNum;
DisplayMz = spectrumSearchResult.HcdSpectrum.IsolationWindow.IsolationWindowTargetMz;
}
else if (spectrumSearchResult.CidSpectrum != null)
{
DisplayScanNum = spectrumSearchResult.CidSpectrum.ScanNum;
DisplayMz = spectrumSearchResult.CidSpectrum.IsolationWindow.IsolationWindowTargetMz;
}
if (spectrumSearchResult.PrecursorSpectrum == null)
{
return;
}
var pearsonCorrelationCalculator = new PearsonCorrelationFitUtil();
PearsonCorrScore = pearsonCorrelationCalculator.GetFitScore(spectrumSearchResult, lipidTarget.Composition);
PearsonCorrScoreMinus1 = pearsonCorrelationCalculator.GetFitMinus1Score(spectrumSearchResult, lipidTarget.Composition);
var cosineCalculator = new CosineFitUtil();
CosineScore = cosineCalculator.GetFitScore(spectrumSearchResult, lipidTarget.Composition);
CosineScoreMinus1 = cosineCalculator.GetFitMinus1Score(spectrumSearchResult, lipidTarget.Composition);
}
public void PearsonCorrelationFileCombiner(string directoryPath)
{
var dirFiles = Directory.GetFiles(directoryPath);
var correlationCalculator = new PearsonCorrelationFitUtil();
var cosineCalculator = new CosineFitUtil();
// Each dictionary corresponds to a dataset, each dictionary key corresponds to the TSV header.
var results = new List <Dictionary <string, List <string> > >();
var headers = new HashSet <string>();
foreach (var pathToResults in dirFiles.Where(path => path.EndsWith(".txt")))
{
var datasetName = Path.GetFileNameWithoutExtension(pathToResults);
var pathToRaw = GetRawFilePath(directoryPath, datasetName);
var rawName = Path.GetFileName(pathToRaw);
if (string.IsNullOrEmpty(pathToRaw))
{
continue;
}
var lcmsRun = PbfLcMsRun.GetLcMsRun(pathToRaw);
var tolerance = new Tolerance(30, ToleranceUnit.Ppm);
using (var reader = new StreamReader(new FileStream(pathToResults, FileMode.Open, FileAccess.Read, FileShare.ReadWrite)))
{
results.Add(new Dictionary <string, List <string> >()); // Add dictionary for new dataset.
var datasetResults = results.Last(); // Results for the current dataset.
var lineCount = 0;
var headerToIndex = new Dictionary <string, int>();
while (!reader.EndOfStream)
{
var line = reader.ReadLine();
if (string.IsNullOrWhiteSpace(line))
{
continue;
}
var pieces = line.Split('\t').ToArray();
if (lineCount++ == 0)
{ // First line
for (var i = 0; i < pieces.Length; i++)
{
var header = pieces[i];
headerToIndex.Add(header, i);
datasetResults.Add(header, new List <string>());
}
datasetResults.Add("Raw File", new List <string>());
datasetResults.Add("Pearson Corr Score", new List <string>());
datasetResults.Add("Pearson Corr M-1 Score", new List <string>());
datasetResults.Add("Cosine Score", new List <string>());
datasetResults.Add("Cosine M-1 Score", new List <string>());
headers.UnionWith(datasetResults.Keys);
continue;
}
var precursor = Convert.ToInt32(pieces[headerToIndex["Precursor Scan"]]);
var commonName = pieces[headerToIndex["Common Name"]];
var adduct = pieces[headerToIndex["Adduct"]];
var spectrum = lcmsRun.GetSpectrum(precursor);
if (spectrum == null)
{
Console.WriteLine("Invalid scan number: {0}", precursor);
continue;
}
var lipid = new Lipid {
AdductFull = adduct, CommonName = commonName
};
var lipidTarget = lipid.CreateLipidTarget();
var spectrumSearchResult = new SpectrumSearchResult(null, null, spectrum, null, null, new Xic(), lcmsRun)
{
PrecursorTolerance = tolerance
};
var pearsonCorrScore = correlationCalculator.GetFitScore(spectrumSearchResult, lipidTarget.Composition);
var pearsonCorrMinus1Score = correlationCalculator.GetFitMinus1Score(spectrumSearchResult, lipidTarget.Composition);
var cosineScore = cosineCalculator.GetFitScore(spectrumSearchResult, lipidTarget.Composition);
var cosineMinus1Score = cosineCalculator.GetFitScore(
spectrumSearchResult,
lipidTarget.Composition);
// Add results to results dictionary.
datasetResults["Raw File"].Add(rawName);
foreach (var header in headerToIndex.Keys)
{
datasetResults[header].Add(pieces[headerToIndex[header]]);
}
datasetResults["Pearson Corr Score"].Add(pearsonCorrScore.ToString());
datasetResults["Pearson Corr M-1 Score"].Add(pearsonCorrMinus1Score.ToString());
datasetResults["Cosine Score"].Add(cosineScore.ToString());
datasetResults["Cosine M-1 Score"].Add(cosineMinus1Score.ToString());
}
}
}
// Write results
var outputFilePath = Path.Combine(directoryPath, "training.tsv");
using (var writer = new StreamWriter(outputFilePath))
{
// Write headers
foreach (var header in headers)
{
writer.Write("{0}\t", header);
}
writer.WriteLine();
// Write data
foreach (var datasetResults in results)
{
var fileLength = datasetResults["Pearson Corr Score"].Count;
for (var i = 0; i < fileLength; i++)
{
foreach (var header in headers)
{
var value = datasetResults.ContainsKey(header) ? datasetResults[header][i] : string.Empty;
writer.Write("{0}\t", value);
}
writer.WriteLine();
}
}
}
}