private List <double> getPeakStatistics()
{
IList <object[]> queryRows;
lock (session)
{
var randomIds = session.CreateQuery("SELECT psm.Id " + viewFilter.GetFilteredQueryString(DataFilter.FromPeptideSpectrumMatch))
.List <long>()
.Shuffle()
.Take(1000)
.OrderBy(o => o);
string randomIdSet = String.Join(",", randomIds.Select(o => o.ToString()).ToArray());
queryRows = session.CreateQuery("SELECT psm.Spectrum.Source.Name, psm.Spectrum, psm, DISTINCT_GROUP_CONCAT(pm.Offset || ':' || mod.MonoMassDelta), psm.Peptide.Sequence " +
"FROM PeptideSpectrumMatch psm " +
"LEFT JOIN psm.Modifications pm " +
"LEFT JOIN pm.Modification mod " +
"WHERE psm.Id IN (" + randomIdSet + ") " +
"GROUP BY psm.Spectrum.id ")
.List <object[]>();
}
var spectrumRows = queryRows.Select(o => new SpectrumRow(o)).OrderBy(o => o.SourceName);
precursorScatterPlot.Clear();
chargeReducedScatterPlot.Clear();
int spectraCount = 0;
string spectrumListFilters = String.Empty;
Invoke(new MethodInvoker(() =>
{
spectrumListFilters = spectrumFiltersTextBox.Text;
zedGraphControl.MasterPane.AxisChange();
zedGraphControl.Refresh();
}));
var points = new PointPairList();
string currentSourceName = null;
string currentSourcePath = null;
msdata.MSData msd = null;
lock (owner)
foreach (var row in spectrumRows)
{
if (row.SourceName != currentSourceName)
{
currentSourceName = row.SourceName;
currentSourcePath = IDPickerForm.LocateSpectrumSource(currentSourceName, session.Connection.GetDataSource());
msd = new pwiz.CLI.msdata.MSDataFile(currentSourcePath);
//var param = session.Query<AnalysisParameter>().Where(o => o.Name == "SpectrumListFilters").Min(o => o.Value);
//string spectrumListFilters = String.IsNullOrEmpty(param) ? String.Empty : param;
SpectrumListFactory.wrap(msd, spectrumListFilters.Split(";".ToCharArray(), StringSplitOptions.RemoveEmptyEntries));
}
string label = String.Format("{0}/{1}\n{2}", row.SourceName, msdata.id.abbreviate(row.Spectrum.NativeID), row.ModifiedSequence);
var spectrumList = msd.run.spectrumList;
++spectraCount;
var pwizPeptide = new proteome.Peptide(row.ModifiedSequence, proteome.ModificationParsing.ModificationParsing_Auto,
proteome.ModificationDelimiter.ModificationDelimiter_Brackets);
var fragmentation = pwizPeptide.fragmentation(true, true);
var pwizSpectrum = spectrumList.spectrum(spectrumList.find(row.Spectrum.NativeID), true);
var pointMap = new seems.PointMap(new ZedGraph.PointPairList(pwizSpectrum.getMZArray().data, pwizSpectrum.getIntensityArray().data));
double tic = pointMap.Values.Sum();
double precursorMz = row.Spectrum.PrecursorMZ;
double chargeReducedPrecursorMz = precursorMz * row.PeptideSpectrumMatch.Charge;
bool plotMatchedPeaks = true;
bool removeMatchedPeaks = false;
double tolerance = 0.03;
seems.PointMap.Enumerator itr;
IonSeries[] ionSeries = Enum.GetValues(typeof(IonSeries)).Cast <IonSeries>().Where(o => o != IonSeries.Count).ToArray();
for (int z = 1; z <= 1; ++z)
{
for (int length = 1, end = pwizPeptide.sequence.Length; length <= end; ++length)
{
string NTermFragment = row.ModifiedSequence.Substring(0, length);
string CTermFragment = row.ModifiedSequence.Substring(row.ModifiedSequence.Length - length);
foreach (IonSeries series in ionSeries)
{
if ((series == IonSeries.c || series == IonSeries.cMinus1 || series == IonSeries.x) &&
length == pwizPeptide.sequence.Length)
{
continue;
}
itr = pointMap.FindNear(fragmentMass(fragmentation, series, length, z), tolerance);
if (itr != null && itr.IsValid)
{
if (plotMatchedPeaks)
{
precursorScatterPlot.AddPoint(new PointPair(itr.Current.Key - precursorMz, itr.Current.Value / tic, (int)series, String.Format("{0} {1}\n{2} {3} {4} {5}", label, precursorMz, NTermFragment, itr.Current.Key, IonSeriesLabels[(int)series], length)));
chargeReducedScatterPlot.AddPoint(new PointPair(itr.Current.Key - chargeReducedPrecursorMz, itr.Current.Value / tic, (int)series, String.Format("{0} {1}\n{2} {3} {4} {5}", label, chargeReducedPrecursorMz, NTermFragment, itr.Current.Key, IonSeriesLabels[(int)series], length)));
}
if (removeMatchedPeaks)
{
pointMap.Remove(itr);
}
}
}
}
}
foreach (var pair in pointMap)
{
precursorScatterPlot.AddPoint(new PointPair(pair.Key - precursorMz, pair.Value / tic, 0, label));
chargeReducedScatterPlot.AddPoint(new PointPair(pair.Key - chargeReducedPrecursorMz, pair.Value / tic, 0, label));
}
if ((spectraCount % 100) == 0)
{
Invoke(new MethodInvoker(() =>
{
zedGraphControl.MasterPane.AxisChange();
zedGraphControl.Refresh();
}));
}
}
Invoke(new MethodInvoker(() =>
{
if (!lockZoomCheckBox.Checked)
{
zedGraphControl.ZoomOutAll(zedGraphControl.GraphPane);
}
zedGraphControl.MasterPane.AxisChange();
zedGraphControl.Refresh();
}));
return(new List <double>()); //percentTicBySpectrumByFragmentType[1];
}
private void getFragmentationStatistics()
{
IList <object[]> queryRows;
lock (session)
{
var randomIds = session.CreateQuery("SELECT psm.Id " + viewFilter.GetFilteredQueryString(DataFilter.FromPeptideSpectrumMatch))
.List <long>()
.Shuffle()
.Take(1000)
.OrderBy(o => o);
string randomIdSet = String.Join(",", randomIds.Select(o => o.ToString()).ToArray());
queryRows = session.CreateQuery("SELECT psm.Spectrum.Source.Name, psm.Spectrum, psm, DISTINCT_GROUP_CONCAT(pm.Offset || ':' || mod.MonoMassDelta), psm.Peptide.Sequence " +
"FROM PeptideSpectrumMatch psm " +
"LEFT JOIN psm.Modifications pm " +
"LEFT JOIN pm.Modification mod " +
"WHERE psm.Id IN (" + randomIdSet + ") " +
"GROUP BY psm.Spectrum.id ")
.List <object[]>();
}
var spectrumRows = queryRows.Select(o => new SpectrumRow(o)).OrderBy(o => o.SourceName);
var percentTicBySpectrumByFragmentType = new List <PointPairList>();
var percentPeakCountBySpectrumByFragmentType = new List <PointPairList>();
var meanMzErrorBySpectrumByFragmentType = new List <PointPairList>();
var percentTicListByFragmentType = new List <List <double> >();
var percentPeakCountListByFragmentType = new List <List <double> >();
var meanMzErrorListByFragmentType = new List <List <double> >();
foreach (var graphControl in graphControls)
{
graphControl.MasterPane.PaneList.ForEach(o => o.CurveList.ForEach(c => c.Clear()));
}
for (int i = 0; i < (int)IonSeries.Count; ++i)
{
percentTicBySpectrumByFragmentType.Add(percentTicGraphControl.MasterPane.PaneList[i + 1].CurveList[3].Points as PointPairList);
percentPeakCountBySpectrumByFragmentType.Add(percentPeakCountGraphControl.MasterPane.PaneList[i + 1].CurveList[3].Points as PointPairList);
meanMzErrorBySpectrumByFragmentType.Add(meanMzErrorGraphControl.MasterPane.PaneList[i + 1].CurveList[3].Points as PointPairList);
percentTicListByFragmentType.Add(new List <double>());
percentPeakCountListByFragmentType.Add(new List <double>());
meanMzErrorListByFragmentType.Add(new List <double>());
}
int spectraCount = 0;
maxPercentTic = 10;
maxPercentPeakCount = 10;
maxMeanMzError = 0.1;
var tolerance = fragmentTolerance;
string spectrumListFilters = String.Empty;
Invoke(new MethodInvoker(() =>
{
tolerance.value = Convert.ToDouble(fragmentToleranceTextBox.Text);
tolerance.units = (MZTolerance.Units)fragmentToleranceUnitsComboBox.SelectedIndex;
meanMzErrorGraphControl.GraphPane.YAxis.Title.Text = "Mean m/z error (" + tolerance.units.ToString() + ")";
spectrumListFilters = spectrumFiltersTextBox.Text;
setAutomaticScales();
}));
var points = new PointPairList();
string currentSourceName = null;
string currentSourcePath = null;
msdata.MSData msd = null;
lock (owner)
foreach (var row in spectrumRows)
{
if (row.SourceName != currentSourceName)
{
currentSourceName = row.SourceName;
currentSourcePath = IDPickerForm.LocateSpectrumSource(currentSourceName, session.Connection.GetDataSource());
if (String.IsNullOrEmpty(currentSourcePath))
{
throw new FileNotFoundException("source file not found");
}
msd = new pwiz.CLI.msdata.MSDataFile(currentSourcePath);
//var param = session.Query<AnalysisParameter>().Where(o => o.Name == "SpectrumListFilters").Min(o => o.Value);
//string spectrumListFilters = String.IsNullOrEmpty(param) ? String.Empty : param;
SpectrumListFactory.wrap(msd, spectrumListFilters.Split(";".ToCharArray(), StringSplitOptions.RemoveEmptyEntries));
}
string spectrumId = String.Format("{0}/{1}", row.SourceName, msdata.id.abbreviate(row.Spectrum.NativeID));
var spectrumList = msd.run.spectrumList;
++spectraCount;
var pwizPeptide = new proteome.Peptide(row.ModifiedSequence, proteome.ModificationParsing.ModificationParsing_Auto,
proteome.ModificationDelimiter.ModificationDelimiter_Brackets);
var fragmentation = pwizPeptide.fragmentation(true, true);
var pwizSpectrum = spectrumList.spectrum(spectrumList.find(row.Spectrum.NativeID), true);
var pointMap = new seems.PointMap(new ZedGraph.PointPairList(pwizSpectrum.getMZArray().data, pwizSpectrum.getIntensityArray().data));
double tic = pointMap.Values.Sum();
var percentTicByFragmentType = new List <double>(Enumerable.Repeat(0.0, (int)IonSeries.Count));
var percentPeakCountByFragmentType = new List <double>(Enumerable.Repeat(0.0, (int)IonSeries.Count));
var matchCountByFragmentType = new List <int>(Enumerable.Repeat(0, (int)IonSeries.Count));
var meanMzErrorByFragmentType = new List <double>(Enumerable.Repeat(Double.NaN, (int)IonSeries.Count));
seems.PointMap.Enumerator itr;
double expected;
IonSeries[] ionSeries = Enum.GetValues(typeof(IonSeries)).Cast <IonSeries>().Where(o => o != IonSeries.Count).ToArray();
for (int z = 1; z <= 1; ++z)
{
for (int length = 1, end = pwizPeptide.sequence.Length; length <= end; ++length)
{
foreach (IonSeries series in ionSeries)
{
if ((series == IonSeries.c || series == IonSeries.x) &&
length == pwizPeptide.sequence.Length)
{
continue;
}
expected = fragmentMass(fragmentation, series, length, z);
itr = pointMap.FindNear(expected, expected - (expected - tolerance));
if (itr != null && itr.IsValid)
{
percentTicByFragmentType[(int)series] += itr.Current.Value;
++percentPeakCountByFragmentType[(int)series];
++matchCountByFragmentType[(int)series];
if (Double.IsNaN(meanMzErrorByFragmentType[(int)series]))
{
meanMzErrorByFragmentType[(int)series] = 0;
}
meanMzErrorByFragmentType[(int)series] += mzError(itr.Current.Key, expected);
}
}
}
}
var rng = new Random();
for (int i = 0; i < percentTicBySpectrumByFragmentType.Count; ++i)
{
// convert sum to mean
if (percentPeakCountByFragmentType[i] > 0)
{
meanMzErrorByFragmentType[i] /= matchCountByFragmentType[i];
}
// convert to percentages
percentTicByFragmentType[i] /= tic / 100;
percentPeakCountByFragmentType[i] /= pointMap.Count / 100.0;
maxPercentTic = Math.Max(maxPercentTic, percentTicByFragmentType[i]);
maxPercentPeakCount = Math.Max(maxPercentPeakCount, percentPeakCountByFragmentType[i]);
double jitter = (rng.NextDouble() - 0.5);
percentTicBySpectrumByFragmentType[i].Add(jitter, percentTicByFragmentType[i], String.Format("{0}: {1:G4}% ({2} matches)", spectrumId, percentTicByFragmentType[i], matchCountByFragmentType[i]));
percentPeakCountBySpectrumByFragmentType[i].Add(jitter, percentPeakCountByFragmentType[i], String.Format("{0}: {1:G4}% ({2} matches)", spectrumId, percentPeakCountByFragmentType[i], matchCountByFragmentType[i]));
percentTicListByFragmentType[i].Add(percentTicByFragmentType[i]);
percentPeakCountListByFragmentType[i].Add(percentPeakCountByFragmentType[i]);
if (!Double.IsNaN(meanMzErrorByFragmentType[i]))
{
maxMeanMzError = Math.Max(maxMeanMzError, Math.Abs(meanMzErrorByFragmentType[i]));
meanMzErrorBySpectrumByFragmentType[i].Add(jitter, meanMzErrorByFragmentType[i], String.Format("{0}: {1:G4}%", spectrumId, meanMzErrorByFragmentType[i]));
meanMzErrorListByFragmentType[i].Add(meanMzErrorByFragmentType[i]);
}
}
if ((spectraCount % 100) == 0)
{
setAutomaticScales();
}
}// for each spectrum row
Invoke(new MethodInvoker(() =>
{
for (int i = 0; i < percentTicBySpectrumByFragmentType.Count; ++i)
{
if (percentTicListByFragmentType[i].Count < 5)
{
continue;
}
percentTicListByFragmentType[i].Sort();
percentPeakCountListByFragmentType[i].Sort();
addSixNumberSummary(percentTicGraphControl.MasterPane.PaneList[i + 1], percentTicListByFragmentType[i]);
addSixNumberSummary(percentPeakCountGraphControl.MasterPane.PaneList[i + 1], percentPeakCountListByFragmentType[i]);
if (meanMzErrorListByFragmentType[i].Count < 5)
{
continue;
}
meanMzErrorListByFragmentType[i].Sort();
addSixNumberSummary(meanMzErrorGraphControl.MasterPane.PaneList[i + 1], meanMzErrorListByFragmentType[i]);
}
}));
}
public SpectrumSource(pwiz.CLI.msdata.MSData msd)
{
Metadata = msd;
}
public static void AddSubsetPeakData (NHibernate.ISession session)
{
session.Transaction.Begin();
session.Clear();
foreach (SpectrumSource source in session.Query<SpectrumSource>())
{
var subsetPeakData = new msdata.MSData();
subsetPeakData.id = subsetPeakData.run.id = source.Name;
var spectrumList = new msdata.SpectrumListSimple();
subsetPeakData.run.spectrumList = spectrumList;
foreach (Spectrum spectrum in source.Spectra.OrderBy(o => o.Index))
{
var spectrumData = new msdata.Spectrum();
spectrumData.id = spectrum.NativeID;
spectrumData.index = spectrum.Index;
spectrumData.setMZIntensityArrays(new List<double>() { 100, 200, 300, 400, 500 },
new List<double>() { 10, 20, 30, 40, 50 });
spectrumList.spectra.Add(spectrumData);
}
session.Evict(source);
var newSource = new SpectrumSource(subsetPeakData)
{
Id = source.Id,
Group = source.Group,
Name = source.Name,
URL = source.URL
};
session.Update(newSource);
}
session.Transaction.Commit();
}