public IEnumerable <Peptide> GenerateAllModificationCombinations()
{
// Get all the modifications that are isotopologues
var isotopologues = GetUniqueModifications <ModificationWithMultiplePossibilitiesCollection>().ToArray();
// Base condition, no more isotopologues to make, so just return
if (isotopologues.Length < 1)
{
yield break;
}
// Grab the the first isotopologue
ModificationWithMultiplePossibilitiesCollection isotopologue = isotopologues[0];
// Loop over each modification in the isotopologue
foreach (OldSchoolModification mod in isotopologue)
{
// Create a clone of the peptide, cloning modifications as well.
Peptide peptide = new Peptide(this);
// Replace the base isotopologue mod with the specific version
peptide.ReplaceModification(isotopologue, mod);
// There were more than one isotopologue, so we must go deeper
if (isotopologues.Length > 1)
{
// Call the same rotuine on the newly generate peptide that has one less isotopologue
foreach (var subpeptide in peptide.GenerateAllModificationCombinations())
{
yield return(subpeptide);
}
}
else
{
// Return this peptide
yield return(peptide);
}
}
}
protected override MetaMorpheusEngineResults RunSpecific()
{
Status("Extracting data points:");
// The final training point list
int numMs1MassChargeCombinationsConsidered = 0;
int numMs1MassChargeCombinationsThatAreIgnoredBecauseOfTooManyPeaks = 0;
int numMs2MassChargeCombinationsConsidered = 0;
int numMs2MassChargeCombinationsThatAreIgnoredBecauseOfTooManyPeaks = 0;
List <LabeledMs1DataPoint> Ms1List = new List <LabeledMs1DataPoint>();
List <LabeledMs2DataPoint> Ms2List = new List <LabeledMs2DataPoint>();
int numIdentifications = goodIdentifications.Count;
// Loop over identifications
HashSet <string> sequences = new HashSet <string>();
object lockObj = new object();
object lockObj2 = new object();
Parallel.ForEach(Partitioner.Create(0, numIdentifications), fff =>
{
for (int matchIndex = fff.Item1; matchIndex < fff.Item2; matchIndex++)
{
PeptideSpectralMatch identification = goodIdentifications[matchIndex];
// Each identification has an MS2 spectrum attached to it.
int ms2scanNumber = identification.ScanNumber;
int peptideCharge = identification.ScanPrecursorCharge;
if (identification.FullSequence == null)
{
continue;
}
var representativeSinglePeptide = identification.CompactPeptides.First().Value.Item2.First();
// Get the peptide, don't forget to add the modifications!!!!
var SequenceWithChemicalFormulas = representativeSinglePeptide.SequenceWithChemicalFormulas;
if (SequenceWithChemicalFormulas == null || representativeSinglePeptide.allModsOneIsNterminus.Any(b => b.Value.neutralLosses.Count != 1 || b.Value.neutralLosses.First() != 0))
{
continue;
}
Proteomics.Peptide coolPeptide = new Proteomics.Peptide(SequenceWithChemicalFormulas);
var ms2tuple = SearchMS2Spectrum(myMsDataFile.GetOneBasedScan(ms2scanNumber) as IMsDataScanWithPrecursor <IMzSpectrum <IMzPeak> >, coolPeptide, peptideCharge, identification);
// If MS2 has low evidence for peptide, skip and go to next one
if (ms2tuple.Item4 < numFragmentsNeededForEveryIdentification)
{
continue;
}
lock (lockObj2)
{
Ms2List.AddRange(ms2tuple.Item1);
numMs2MassChargeCombinationsConsidered += ms2tuple.Item2;
numMs2MassChargeCombinationsThatAreIgnoredBecauseOfTooManyPeaks += ms2tuple.Item3;
if (sequences.Contains(identification.FullSequence))
{
continue; // Do not search same sequence multiple times in MS1 scans
}
sequences.Add(identification.FullSequence);
}
// Calculate isotopic distribution of the full peptide
var dist = IsotopicDistribution.GetDistribution(coolPeptide.GetChemicalFormula(), fineResolutionForIsotopeDistCalculation, 0.001);
double[] theoreticalMasses = dist.Masses.ToArray();
double[] theoreticalIntensities = dist.Intensities.ToArray();
Array.Sort(theoreticalIntensities, theoreticalMasses, Comparer <double> .Create((x, y) => y.CompareTo(x)));
var ms1tupleBack = SearchMS1Spectra(theoreticalMasses, theoreticalIntensities, ms2scanNumber, -1, peptideCharge, identification);
var ms1tupleForward = SearchMS1Spectra(theoreticalMasses, theoreticalIntensities, ms2scanNumber, 1, peptideCharge, identification);
lock (lockObj)
{
Ms1List.AddRange(ms1tupleBack.Item1);
numMs1MassChargeCombinationsConsidered += ms1tupleBack.Item2;
numMs1MassChargeCombinationsThatAreIgnoredBecauseOfTooManyPeaks += ms1tupleBack.Item3;
Ms1List.AddRange(ms1tupleForward.Item1);
numMs1MassChargeCombinationsConsidered += ms1tupleForward.Item2;
numMs1MassChargeCombinationsThatAreIgnoredBecauseOfTooManyPeaks += ms1tupleForward.Item3;
}
}
});
return(new DataPointAquisitionResults(this,
Ms1List,
Ms2List,
numMs1MassChargeCombinationsConsidered,
numMs1MassChargeCombinationsThatAreIgnoredBecauseOfTooManyPeaks,
numMs2MassChargeCombinationsConsidered,
numMs2MassChargeCombinationsThatAreIgnoredBecauseOfTooManyPeaks
));
}
private (List <LabeledMs2DataPoint>, int, int, int) SearchMS2Spectrum(IMsDataScanWithPrecursor <IMzSpectrum <IMzPeak> > ms2DataScan, Proteomics.Peptide peptide, int peptideCharge, PeptideSpectralMatch identification)
{
List <LabeledMs2DataPoint> result = new List <LabeledMs2DataPoint>();
int numMs2MassChargeCombinationsConsidered = 0;
int numMs2MassChargeCombinationsThatAreIgnoredBecauseOfTooManyPeaks = 0;
int numFragmentsIdentified = 0;
if (ms2DataScan.MassSpectrum.Size == 0)
{
return(result, numMs2MassChargeCombinationsConsidered, numMs2MassChargeCombinationsThatAreIgnoredBecauseOfTooManyPeaks, numFragmentsIdentified);
}
// Key: mz value, Value: error
var addedPeaks = new Dictionary <double, double>();
var countForThisMS2 = 0;
var countForThisMS2a = 0;
var scanWindowRange = ms2DataScan.ScanWindowRange;
IHasChemicalFormula[] fragmentList = peptide.Fragment(fragmentTypesForCalibration, true).OfType <IHasChemicalFormula>().ToArray();
foreach (var fragment in fragmentList)
{
bool fragmentIdentified = false;
bool computedIsotopologues = false;
double[] masses = new double[0];
double[] intensities = new double[0];
// First look for monoisotopic masses, do not compute distribution spectrum!
for (int chargeToLookAt = 1; chargeToLookAt <= peptideCharge; chargeToLookAt++)
{
var monoisotopicMZ = fragment.MonoisotopicMass.ToMz(chargeToLookAt);
if (monoisotopicMZ > scanWindowRange.Maximum)
{
continue;
}
if (monoisotopicMZ < scanWindowRange.Minimum)
{
break;
}
var closestPeakMZ = ms2DataScan.MassSpectrum.GetClosestPeakXvalue(monoisotopicMZ);
if (mzToleranceForMs2Search.Within(closestPeakMZ.Value, monoisotopicMZ) && !computedIsotopologues)
{
var dist = IsotopicDistribution.GetDistribution(fragment.ThisChemicalFormula, fineResolutionForIsotopeDistCalculation, 0.001);
masses = dist.Masses.ToArray();
intensities = dist.Intensities.ToArray();
Array.Sort(intensities, masses, Comparer <double> .Create((x, y) => y.CompareTo(x)));
computedIsotopologues = true;
break;
}
}
if (computedIsotopologues)
{
bool startingToAdd = false;
for (int chargeToLookAt = 1; chargeToLookAt <= peptideCharge; chargeToLookAt++)
{
if (masses.First().ToMz(chargeToLookAt) > scanWindowRange.Maximum)
{
continue;
}
if (masses.Last().ToMz(chargeToLookAt) < scanWindowRange.Minimum)
{
break;
}
var trainingPointsToAverage = new List <LabeledMs2DataPoint>();
foreach (double a in masses)
{
double theMZ = a.ToMz(chargeToLookAt);
var npwr = ms2DataScan.MassSpectrum.NumPeaksWithinRange(mzToleranceForMs2Search.GetMinimumValue(theMZ), mzToleranceForMs2Search.GetMaximumValue(theMZ));
if (npwr == 0)
{
break;
}
numMs2MassChargeCombinationsConsidered++;
if (npwr > 1)
{
numMs2MassChargeCombinationsThatAreIgnoredBecauseOfTooManyPeaks++;
continue;
}
var closestPeakIndex = ms2DataScan.MassSpectrum.GetClosestPeakIndex(theMZ);
var closestPeakMZ = ms2DataScan.MassSpectrum.XArray[closestPeakIndex.Value];
if (!addedPeaks.ContainsKey(closestPeakMZ))
{
addedPeaks.Add(closestPeakMZ, Math.Abs(closestPeakMZ - theMZ));
trainingPointsToAverage.Add(new LabeledMs2DataPoint(closestPeakMZ, double.NaN, double.NaN, double.NaN, Math.Log(ms2DataScan.MassSpectrum.YArray[closestPeakIndex.Value]), theMZ, null));
}
}
// If started adding and suddnely stopped, go to next one, no need to look at higher charges
if (trainingPointsToAverage.Count == 0 && startingToAdd)
{
break;
}
if (trainingPointsToAverage.Count < Math.Min(minMS2isotopicPeaksNeededForConfirmedIdentification, intensities.Count()))
{
}
else
{
startingToAdd = true;
if (!fragmentIdentified)
{
fragmentIdentified = true;
numFragmentsIdentified += 1;
}
countForThisMS2 += trainingPointsToAverage.Count;
countForThisMS2a++;
result.Add(new LabeledMs2DataPoint(trainingPointsToAverage.Select(b => b.mz).Average(),
ms2DataScan.RetentionTime,
Math.Log(ms2DataScan.TotalIonCurrent),
ms2DataScan.InjectionTime.HasValue ? Math.Log(ms2DataScan.InjectionTime.Value) : double.NaN,
trainingPointsToAverage.Select(b => b.logIntensity).Average(),
trainingPointsToAverage.Select(b => b.expectedMZ).Average(),
identification));
}
}
}
}
return(result, numMs2MassChargeCombinationsConsidered, numMs2MassChargeCombinationsThatAreIgnoredBecauseOfTooManyPeaks, numFragmentsIdentified);
}
public static void TestCoIsolation()
{
CommonParameters CommonParameters = new CommonParameters
{
DigestionParams = new DigestionParams
{
Protease = new Protease("Custom Protease", new List <string> {
"K"
}, new List <string>(), TerminusType.C, CleavageSpecificity.Full, null, null, null),
MinPeptideLength = null,
},
ConserveMemory = false,
ScoreCutoff = 1,
DeconvolutionIntensityRatio = 50
};
var variableModifications = new List <ModificationWithMass>();
var fixedModifications = new List <ModificationWithMass>();
var proteinList = new List <Protein> {
new Protein("MNNNKNDNK", null)
};
var searchModes = new SinglePpmAroundZeroSearchMode(5);
Proteomics.Peptide pep1 = new Proteomics.Peptide("NNNK");
Proteomics.Peptide pep2 = new Proteomics.Peptide("NDNK");
var dist1 = IsotopicDistribution.GetDistribution(pep1.GetChemicalFormula(), 0.1, 0.01);
var dist2 = IsotopicDistribution.GetDistribution(pep2.GetChemicalFormula(), 0.1, 0.01);
IMzmlScan[] Scans = new IMzmlScan[2];
double[] ms1intensities = new double[] { 0.8, 0.8, 0.2, 0.02, 0.2, 0.02 };
double[] ms1mzs = dist1.Masses.Concat(dist2.Masses).OrderBy(b => b).Select(b => b.ToMz(1)).ToArray();
double selectedIonMz = ms1mzs[1];
MzmlMzSpectrum MS1 = new MzmlMzSpectrum(ms1mzs, ms1intensities, false);
Scans[0] = new MzmlScan(1, MS1, 1, false, Polarity.Positive, 1.0, new MzRange(300, 2000), "first spectrum", MZAnalyzerType.Unknown, MS1.SumOfAllY, null, "scan=1");
double[] ms2intensities = new double[] { 1, 1, 1, 1, 1 };
double[] ms2mzs = new double[] { 146.106.ToMz(1), 228.086.ToMz(1), 229.07.ToMz(1), 260.148.ToMz(1), 342.129.ToMz(1) };
MzmlMzSpectrum MS2 = new MzmlMzSpectrum(ms2mzs, ms2intensities, false);
double isolationMZ = selectedIonMz;
Scans[1] = new MzmlScanWithPrecursor(2, MS2, 2, false, Polarity.Positive, 2.0, new MzRange(100, 1500), "second spectrum", MZAnalyzerType.Unknown, MS2.SumOfAllY, selectedIonMz, null, null, isolationMZ, 2.5, DissociationType.HCD, 1, null, null, "scan=2");
var myMsDataFile = new FakeMsDataFile(Scans);
bool DoPrecursorDeconvolution = true;
bool UseProvidedPrecursorInfo = true;
double DeconvolutionIntensityRatio = 50;
int DeconvolutionMaxAssumedChargeState = 10;
Tolerance DeconvolutionMassTolerance = new PpmTolerance(5);
var listOfSortedms2Scans = MetaMorpheusTask.GetMs2Scans(myMsDataFile, null, DoPrecursorDeconvolution, UseProvidedPrecursorInfo, DeconvolutionIntensityRatio, DeconvolutionMaxAssumedChargeState, DeconvolutionMassTolerance).OrderBy(b => b.PrecursorMass).ToArray();
PeptideSpectralMatch[] allPsmsArray = new PeptideSpectralMatch[listOfSortedms2Scans.Length];
List <ProductType> lp = new List <ProductType> {
ProductType.B, ProductType.Y
};
new ClassicSearchEngine(allPsmsArray, listOfSortedms2Scans, variableModifications, fixedModifications, proteinList, lp, searchModes, false, CommonParameters, CommonParameters.ProductMassTolerance, new List <string>()).Run();
// Two matches for this single scan! Corresponding to two co-isolated masses
Assert.AreEqual(2, allPsmsArray.Length);
Assert.IsTrue(allPsmsArray[0].Score > 1);
Assert.AreEqual(2, allPsmsArray[0].ScanNumber);
var ojdfkj = (SequencesToActualProteinPeptidesEngineResults) new SequencesToActualProteinPeptidesEngine(new List <PeptideSpectralMatch> {
allPsmsArray[0], allPsmsArray[1]
}, proteinList, fixedModifications, variableModifications, lp, new List <IDigestionParams> {
CommonParameters.DigestionParams
}, CommonParameters.ReportAllAmbiguity, new List <string>()).Run();
foreach (var huh in allPsmsArray)
{
if (huh != null)
{
huh.MatchToProteinLinkedPeptides(ojdfkj.CompactPeptideToProteinPeptideMatching);
}
}
Assert.AreEqual("NNNK", allPsmsArray[0].BaseSequence);
Assert.AreEqual("NDNK", allPsmsArray[1].BaseSequence);
}