}//*/
public IEnumerable <ProductMatch> GetProductMZs(DBOptions options, GraphML_List <MsMsPeak> peaks)//, List<double> theoretical_product_mzs = null)
{
// speed optimizations
//double[] experimental_masses = Query.spectrum.Masses;
//GraphML_List<MSPeak> peaks = Query.spectrum.Peaks;
int num_experimental_peaks = peaks.Count;
TotalTheoreticalProducts = 0;
TotalWeightedProducts = 0;
//New version that should include charged ions
//foreach (ProductMatch matchTheo in AllFragmentSearch.ComputeAllFragments(Peptide, Query.precursor.Charge, options))
foreach (ProductMatch matchTheo in options.fragments.ComputeFragmentsFast(Peptide.GetMasses(), Query.precursor.Charge, options))
//foreach (ProductMatch matchTheo in options.fragments.ComputeFragments(Peptide, Query.precursor.Charge, options))
{
TotalTheoreticalProducts++;
TotalWeightedProducts += matchTheo.weight;//++;
//TODO test what is most common between charged ions, and uncharged ions
//for (int charge = 1; charge <= Query.precursor.Charge; charge++)
//for (int charge = Query.precursor.Charge; charge >= 1; charge--)
//{
double massDiff = options.productMassTolerance.Value;
double bestMz = -1;
double bestInt = 0;
foreach (int index in Query.spectrum.GetIndexOfMZInRange(matchTheo.theoMz, options.productMassTolerance))
{
if (peaks[index].Charge <= 0 || peaks[index].Charge == matchTheo.charge)
{
double diff = Numerics.CalculateMassError(peaks[index].MZ, matchTheo.theoMz, options.productMassTolerance.Units);
//double diff = matchTheo.theoMz - peaks[index].MZ;// experimental_masses[index];//TODO DALTON ONLY : add product mass tolerance unit test
if (Math.Abs(diff) < options.productMassTolerance.Value)
{
if (Math.Abs(diff) < Math.Abs(massDiff))//TODO Priority to intensity, or precision?
{
massDiff = diff;
bestMz = peaks[index].MZ;
}
//if (peaks[index].Intensity > bestInt)
bestInt += peaks[index].Intensity;
}
}
}
if (bestMz >= 0)
{
ProductMatch pMatch = new ProductMatch();
pMatch.weight = matchTheo.weight;
pMatch.theoMz = matchTheo.theoMz; // Utilities.MZFromMzSingleCharge(theoMass, charge);
pMatch.obsMz = bestMz; // experimental_masses[bestIndex];
pMatch.mass_diff = massDiff;
pMatch.obsIntensity = bestInt; // Intensities[bestIndex];
pMatch.charge = matchTheo.charge; // peaks[bestIndex].Charge;
pMatch.Fragment = matchTheo.Fragment;
pMatch.fragmentPos = matchTheo.fragmentPos;
pMatch.normalizedIntensity = pMatch.obsIntensity / (Query.spectrum.InjectionTime * Query.spectrum.PrecursorIntensityPerMilliSecond);
yield return(pMatch);
//break;
}
}
}
public Precursor()
{
this.psms_AllPossibilities = new PeptideSpectrumMatches(); // GraphML_List<PeptideSpectrumMatch>();
this.psms = new PeptideSpectrumMatches(); // GraphML_List<PeptideSpectrumMatch>();
Isotopes = new GraphML_List <Precursor>();
OtherCharges = new GraphML_List <Precursor>();
}
private static GraphML_List <MsMsPeak> AssignChargeStatesbkp(GraphML_List <MsMsPeak> peaks, int maxCharge, MassTolerance isotopicMzTolerance)
{
GraphML_List <MsMsPeak> new_peaks = new GraphML_List <MsMsPeak>();
for (int i = 0; i < peaks.Count - 1; i++)
{
int j = i + 1;
List <int> charges = new List <int>();
while (j < peaks.Count)
{
if (peaks[j].MZ > (peaks[i].MZ + Constants.C12_C13_MASS_DIFFERENCE) + isotopicMzTolerance)
{
break;
}
for (int c = maxCharge; c >= 1; c--)
{
if (Math.Abs(Numerics.CalculateMassError(peaks[j].MZ, peaks[i].MZ + Constants.C12_C13_MASS_DIFFERENCE / (double)c, isotopicMzTolerance.Units)) <= isotopicMzTolerance.Value)
{
new_peaks.Add(new MsMsPeak(peaks[i].MZ, peaks[i].Intensity, c));
charges.Add(c);
}
}
j++;
}
if (charges.Count == 0)
{
new_peaks.Add(new MsMsPeak(peaks[i].MZ, peaks[i].Intensity, 0));
}
}
return(new_peaks);
}
public Precursor(Track track, int charge, Sample entry, double massShift = 0, GraphML_List <Precursor> isotopes = null)//double mz, double intensity, int charge = -1, double rt = 0, double massShift = 0, List<Precursor> isotopes = null)
{
INDEX = COMPTEUR++;
this.Track = track;
this.Charge = charge;
this.sample = entry;
this.psms_AllPossibilities = new PeptideSpectrumMatches(); // GraphML_List<PeptideSpectrumMatch>();
this.psms = new PeptideSpectrumMatches(); // GraphML_List<PeptideSpectrumMatch>();
// this.Mz = mz;
// this.Intensity = intensity;
// this.Charge = charge;
this.Mass = Numerics.MassFromMZ(track.MZ, charge);
// this.Rt = rt;
MassShift = massShift;
if (isotopes == null)
{
Isotopes = new GraphML_List <Precursor>();
}
else
{
Isotopes = isotopes;
}
OtherCharges = new GraphML_List <Precursor>();
}
public clCondition(int nbReplicates)
{
replicates = new GraphML_List <clReplicate>();
for (int i = 0; i < nbReplicates; i++)
{
replicates.Add(null);
}
}
public Cluster(Dictionary <int, List <int> > samples)
{
this.conditions = new GraphML_List <clCondition>();
foreach (int c in samples.Keys)
{
this.conditions.Add(null);
}
this.samples = samples;
}
private static GraphML_List <MsMsPeak> FilterPeaks(GraphML_List <MsMsPeak> peaks, int maximumNumberOfPeaks)
{
GraphML_List <MsMsPeak> filtered_peaks = new GraphML_List <MsMsPeak>(peaks);
if (maximumNumberOfPeaks > 0 && filtered_peaks.Count > maximumNumberOfPeaks)
{
filtered_peaks.Sort(MsMsPeak.DescendingIntensityComparison);
filtered_peaks.RemoveRange(maximumNumberOfPeaks, filtered_peaks.Count - maximumNumberOfPeaks);
}
return(filtered_peaks);
}
public Result()
{
queries = new Queries();
precursors = new Precursors();
matchedPrecursors = new Precursors();
clusters = new GraphML_List <Cluster>();
peptides = new PeptideMatches(new PeptideMatch[0]);
peptideSequences = new PeptideMatches(new PeptideMatch[0]);
proteins = new ProteinGroupMatches();
dbOptions = new DBOptions("");
samples = new Samples();
}
public static GraphML_List <Precursor> GetIsotopes(Track track, DBOptions dbOptions, Tracks listTracks, Sample entry)
{
double isotopicMzTolerance = dbOptions.precursorMassTolerance.Value * 0.5;
if (dbOptions.precursorMassTolerance.Units == MassToleranceUnits.ppm)
{
isotopicMzTolerance = (isotopicMzTolerance / 1e6) * track.MZ;
}
GraphML_List <Precursor> bestIsotopes = new GraphML_List <Precursor>();
for (int charge = dbOptions.MaximumPrecursorChargeState; charge >= dbOptions.MinimumPrecursorChargeState; charge--)
{
GraphML_List <Precursor> isotopes = new GraphML_List <Precursor>();
for (int nbIsotope = 1; nbIsotope < 5; nbIsotope++)
{
double bestDeltaMz = isotopicMzTolerance;
Track bestTrack = null;
double massShift = Numerics.IsotopicMassShift(nbIsotope, charge);
double mzIsotope = track.MZ + massShift;
foreach (Track trackToTest in listTracks.GetTracksInMzRange(mzIsotope, isotopicMzTolerance))
{
if (trackToTest.RT >= track.RT_Min &&
trackToTest.RT <= track.RT_Max)
{
double mzDiff = Math.Abs(mzIsotope - trackToTest.MZ);
if (mzDiff < bestDeltaMz)//TODO Is the best isotope the most precise one or the most intense?? Use a scoring function!
{
bestDeltaMz = mzDiff;
bestTrack = trackToTest;
}
}
}
if (bestTrack != null)
{
isotopes.Add(new Precursor(bestTrack, charge, entry, Constants.C12_C13_MASS_DIFFERENCE * nbIsotope, null));
}
else
{
break;
}
}
if (isotopes.Count > bestIsotopes.Count)//TODO Best way to compare isotope potentials? Number of isotopes? Delta Mass? Intensity ratios?
{
bestIsotopes = isotopes;
}
}
return(bestIsotopes);
}
/// <summary>
/// Parameter less constructor used for save states
/// </summary>
public DBOptions(IConSol console = null)
{
fixedModifications = new GraphML_List <Modification>();
variableModifications = new GraphML_List <Modification>();
fragments = new Fragments();
if (console == null)
{
ConSole = new ConSolCommandLine();
}
else
{
ConSole = console;
}
}
public static Spectra Import(string filenameMSMS, string filenameTracks, DBOptions dbOptions)
{
Spectra spectra = new Spectra();
vsCSV csv = new vsCSV(filenameMSMS);
if (csv.LINES_LIST.Count == 0 || csv.LINES_LIST[0].CompareTo(ProductSpectrum.TITLE) != 0)
{
return(null);
}
for (int i = 1; i < csv.LINES_LIST.Count; i++)
{
string[] splits = csv.LINES_LIST[i].Split(vsCSV._Generic_Separator);
double mz = double.Parse(splits[3]);
int charge = int.Parse(splits[5]);
int nbPeaks = int.Parse(splits[9]);
GraphML_List <MsMsPeak> peaks = new GraphML_List <MsMsPeak>(nbPeaks);
i++;
for (int j = 0; j < nbPeaks; i++, j++)
{
try
{
string[] splitPeaks = csv.LINES_LIST[i].Split('\t');
if (splitPeaks.Length > 2)
{
peaks.Add(new MsMsPeak(double.Parse(splitPeaks[0]), double.Parse(splitPeaks[1]), int.Parse(splitPeaks[2])));
}
else
{
peaks.Add(new MsMsPeak(double.Parse(splitPeaks[0]), double.Parse(splitPeaks[1]), 0));
}
}
catch (Exception)
{
dbOptions.ConSole.WriteLine("Error parsing line : " + csv.LINES_LIST[i]);
}
}
spectra.AddMSMS(new ProductSpectrum(int.Parse(splits[0]), double.Parse(splits[1]), splits[2], mz, double.Parse(splits[4]), charge, Proteomics.Utilities.Numerics.MassFromMZ(mz, charge), peaks, double.Parse(splits[8]), double.Parse(splits[10]), double.Parse(splits[11])));
}
if (!string.IsNullOrEmpty(filenameTracks))
{
spectra.tracks = Tracks.Import(filenameTracks, dbOptions);
}
return(spectra);
}
public GraphML_List <Cluster> Search(Precursors precursors, bool runCluster)
{
options.ConSole.WriteLine("Grouping precursors based on common features...");
precursors.Sort(Precursor.CompareProbabilityScore);
GraphML_List <Cluster> clusters = new GraphML_List <Cluster>();
bool[] done = new bool[precursors.Count];
for (int i = 0; i < done.Length; i++)
{
done[i] = false;
}
//Step 1 : Regroup psms based on mz/rt/intensity/Sequence proximity score (ProteoProfile Code)
for (int i = 0; i < precursors.Count; i++)
{
if (!done[i])
{
Cluster group = new Cluster(samples);
group.Add(precursors[i]);
if (runCluster)
{
for (int j = i + 1; j < precursors.Count; j++)
{
if (!done[j] && precursors[i].sample != precursors[j].sample)
{
double score = Score(precursors[i], precursors[j]);
//TODO Implement ProteoProfile Clustering algorithm, or anything on the litterature, as long as its backed by the scoring function
if (score > 0.75)//TODO Should we put a threshold here? Can it be computed dynamically?
{
group.Add(precursors[j]);
done[j] = true;
}
}
}
}
clusters.Add(group);
}
}
options.ConSole.WriteLine("Created " + clusters.Count + " clusters");
return(clusters);
//TODO I should not use psms in more than one cluster...
}
public static GraphML_List <Precursor> GetOtherCharges(Precursor precursor, DBOptions dbOptions, Tracks listTracks, Sample entry)
{
GraphML_List <Precursor> otherPrecursor = new GraphML_List <Precursor>();
for (int charge = dbOptions.MaximumPrecursorChargeState; charge >= 1; charge--)
{
if (charge != precursor.Charge)
{
double aimedMZ = Numerics.MZFromMass(precursor.Mass, charge);
double chargeMzTolerance = dbOptions.precursorMassTolerance.Value * 0.5;
if (dbOptions.precursorMassTolerance.Units == MassToleranceUnits.ppm)
{
chargeMzTolerance = (chargeMzTolerance / 1e6) * aimedMZ;
}
double bestDeltaMz = chargeMzTolerance;
Track bestTrack = null;
foreach (Track trackToTest in listTracks.GetTracksInMzRange(aimedMZ, chargeMzTolerance))
{
if (trackToTest.RT >= precursor.Track.RT_Min &&
trackToTest.RT <= precursor.Track.RT_Max)
{
double mzDiff = Math.Abs(aimedMZ - trackToTest.MZ);
if (mzDiff < bestDeltaMz)//TODO Is the best isotope the most precise one or the closest in intensity?? Use a scoring function to get both!
{
bestDeltaMz = mzDiff;
bestTrack = trackToTest;
}
}
}
if (bestTrack != null)
{
otherPrecursor.Add(new Precursor(bestTrack, charge, precursor.sample, 0, GetIsotopes(bestTrack, dbOptions, listTracks, entry)));
}
}
}
return(otherPrecursor);
}
public void Initialize(DBOptions options, IEnumerable <ProductMatch> productMZs)
{
//List<double> theoretical_product_mz = Peptide.CalculateAllProductMz(PRODUCT_TYPES[Query.spectrum.FragmentationMethod], Query.precursor);
//TotalProducts = theoretical_product_mz.Count;
MatchingProducts = 0;
MatchingIntensity = 0.0;
double cumulDiff = 0;
ProductScore = 0;
GraphML_List <ProductMatch> cumulMatch = new GraphML_List <ProductMatch>();
MatchingWeightedProducts = 0;
foreach (ProductMatch match in productMZs)
{
MatchingProducts++;
MatchingWeightedProducts += match.weight;
MatchingIntensity += match.obsIntensity;
cumulDiff += Math.Abs(match.mass_diff) * match.weight;
cumulMatch.Add(match);
if (match.obsIntensity > highestFragmentIntensity)
{
highestFragmentIntensity = match.obsIntensity;
}
//options.ConSole.WriteLine("fragment intensity higher than most intense fragment ... should not happen!");
}
AllProductMatches = cumulMatch;
MatchingProductsFraction = (double)MatchingWeightedProducts / (double)TotalWeightedProducts;
MatchingIntensityFraction = MatchingIntensity / (double)(highestFragmentIntensity * TotalTheoreticalProducts);
if (MatchingIntensityFraction > 1)
{
MatchingIntensityFraction = 1.0;
}
ProductScore = 1.0 - (cumulDiff / (double)(MatchingProducts * options.productMassTolerance.Value));
IntensityScore = MatchingIntensityFraction / (double)Query.spectrum.Peaks.Count;
MorpheusScore = MatchingProducts + MatchingIntensityFraction;
FragmentScore = ComputeFragmentScore();
}
private static GraphML_List <MsMsPeak> Deisotopebkp(GraphML_List <MsMsPeak> peaks, int maxCharge, MassTolerance isotopicMzTolerance)
{
GraphML_List <MsMsPeak> new_peaks = new GraphML_List <MsMsPeak>(peaks);
peaks.Sort(MsMsPeak.AscendingMzComparison);
for (int lowMassIndex = 0; lowMassIndex < new_peaks.Count - 1; lowMassIndex++)
{
if (new_peaks[lowMassIndex].Charge > 0)
{
int toRemove = -1;
double bestMassError = isotopicMzTolerance.Value;
double aim = Numerics.IsotopicMassShift(1, new_peaks[lowMassIndex].Charge) + new_peaks[lowMassIndex].MZ;
int potentialIsotopeIndex = lowMassIndex + 1;
while (potentialIsotopeIndex < new_peaks.Count && new_peaks[potentialIsotopeIndex].MZ < aim + bestMassError)
{
if (new_peaks[lowMassIndex].Intensity > new_peaks[potentialIsotopeIndex].Intensity)
{
double massError = Math.Abs(Numerics.CalculateMassError(new_peaks[potentialIsotopeIndex].MZ, aim, isotopicMzTolerance.Units));
if (massError < bestMassError)
{
bestMassError = massError;
toRemove = potentialIsotopeIndex;
}
}
potentialIsotopeIndex++;
}
if (toRemove > 0)
{
new_peaks[lowMassIndex].Intensity += new_peaks[toRemove].Intensity;
new_peaks.RemoveAt(toRemove);
}
}
}
return(new_peaks);
}
public DBOptions(string fasta, IConSol console = null)
{
if (console == null)
{
ConSole = new ConSolCommandLine();
}
else
{
ConSole = console;
}
//Create with default values
this.DecoyFusion = true;
this.FastaDatabaseFilepath = fasta;
this.MaximumPeptideMass = 10000;
ProteaseDictionary proteases = ProteaseDictionary.Instance;
this.DigestionEnzyme = proteases["no enzyme"]; // proteases["trypsin (no proline rule)"];
this.NoEnzymeSearch = true;
this.ToleratedMissedCleavages = 100; // 3;//determines the length of peptides with no-enzyme option
this.initiatorMethionineBehavior = InitiatorMethionineBehavior.Variable;
this.fixedModifications = new GraphML_List <Modification>();
this.variableModifications = new GraphML_List <Modification>();
this.maximumVariableModificationIsoforms = 1024;
this.MinimumPrecursorChargeState = 1;
this.MaximumPrecursorChargeState = 4;
this.MaximumNumberOfFragmentsPerSpectrum = 400;
//TODO Add precision to the precursor by reading MS part of file
this.precursorMassTolerance = new MassTolerance(0.005, MassToleranceUnits.Da);//2.1
//TODO Add precision to the product masses by reading corresponding MS part of raw file
this.productMassTolerance = new MassTolerance(0.005, MassToleranceUnits.Da);
this.PSMFalseDiscoveryRate = 0.25; // 0.05;
this.OutputFolder = @"C:\_IRIC\DATA\Test2"; //C:\Documents and Settings\ProteoAdmin\Desktop\AEffacer\Morpheus\Output";
this.MinimumPSMScore = 0.0001;
}
public clCondition()
{
replicates = new GraphML_List <clReplicate>();
}
public clReplicate()
{
precursors = new GraphML_List <Precursor>();
}
public PeptideSpectrumMatch()
{
AllProductMatches = new GraphML_List <ProductMatch>();
}
public Cluster()
{
conditions = new GraphML_List <clCondition>();
}
public void GenerateQueries(Sample entry, Spectra spectra, Tracks tracks)//, double mz, double rt, double intensity)
{
Dictionary <Track, Precursor> Tracks = new Dictionary <Track, Precursor>();
Dictionary <Track, Precursor> Isotopes = new Dictionary <Track, Precursor>();
//Create one query per Spectrum-Precursor duo, including Isotopes in the process to ease search
//For further analysis, maintain a list of precursors (excluding isotopes)
int nbMissedTrack = 0;
//vsSDF sdf = entry.GetSDF();// Samples.LoadSDF(entry);
//tracks.PrepareRtSort();
//sdf.TRACKS_LIST.PrepareRtSort();
spectra.Sort(ProductSpectrum.AscendingPrecursorMassComparison);
foreach (ProductSpectrum spectrum in spectra)
{
NbSpectrum++;
double intensityCumul = 0.0;
bool foundCharge = false;
Track closestTrack = null;
List <Query> newQueries = new List <Query>();
//TODO No threshold on sdf files, and preferably a C# routine that does what MassSense do
foreach (Track track in tracks.GetTracksInMzRange(spectrum.PrecursorMZ, spectrum.IsolationWindow * dbOptions.EffectiveIsolationWindowRatio))//TODO Optimize this value
{
Precursor prec = null;
if (track.RT_Min <= spectrum.RetentionTimeInMin &&
track.RT_Max >= spectrum.RetentionTimeInMin)
{
if (closestTrack == null || Math.Abs(track.MZ - spectrum.PrecursorMZ) < Math.Abs(closestTrack.MZ - spectrum.PrecursorMZ))
{
closestTrack = track;
}
if (Isotopes.ContainsKey(track))
{
break;
}
if (Tracks.ContainsKey(track))
{
prec = Tracks[track];
}
else
{
GraphML_List <Precursor> isotopes = GetIsotopes(track, dbOptions, tracks, entry);
if (isotopes.Count > 0)
{
prec = new Precursor(track, isotopes[0].Charge, entry, 0.0, isotopes);
Tracks.Add(track, prec);
prec.OtherCharges = GetOtherCharges(prec, dbOptions, tracks, entry);
foreach (Precursor isotope in prec.Isotopes)
{
if (!Isotopes.ContainsKey(isotope.Track))
{
Isotopes.Add(isotope.Track, isotope);
}
}
}
}
if (prec != null)
{
intensityCumul += track.INTENSITY;
newQueries.Add(new Query(dbOptions, entry, spectrum, prec, NbSpectrum));
if (prec.Charge == spectrum.PrecursorCharge)
{
foundCharge = true;
}
}
}
}
if (!foundCharge)
{
/*if (closestTrack != null && Tracks.ContainsKey(closestTrack) && Math.Abs(Numerics.CalculateMassError(closestTrack.MZ, spectrum.PrecursorMZ, dbOptions.precursorMassTolerance.Units)) < dbOptions.precursorMassTolerance.Value)
* {
* if(closestTrack.RT_Min > (float)(spectrum.RetentionTimeInMin - dbOptions.ComputedRetentionTimeDiff))
* closestTrack.RT_Min = (float)(spectrum.RetentionTimeInMin - dbOptions.ComputedRetentionTimeDiff);
* if (closestTrack.RT_Max < (float)(spectrum.RetentionTimeInMin + dbOptions.ComputedRetentionTimeDiff))
* closestTrack.RT_Max = (float)(spectrum.RetentionTimeInMin + dbOptions.ComputedRetentionTimeDiff);
* if (closestTrack.INTENSITY < spectrum.PrecursorIntensity)
* closestTrack.INTENSITY = spectrum.PrecursorIntensity;
*
* Precursor prec = Tracks[closestTrack];
* if (prec.Charge == spectrum.PrecursorCharge)
* {
* Add(new Query(dbOptions, entry, spectrum, prec, NbSpectrum));
* }
* else
* {
* Precursor newPrec = new Precursor(closestTrack, spectrum.PrecursorCharge, entry);
* Add(new Query(dbOptions, entry, spectrum, newPrec, NbSpectrum));
* }
* }
* else//*/
{
nbMissedTrack++;
closestTrack = new Track((float)spectrum.PrecursorMZ, (float)spectrum.RetentionTimeInMin, spectrum.PrecursorIntensity,
(float)(spectrum.RetentionTimeInMin - dbOptions.ComputedRetentionTimeDiff), (float)(spectrum.RetentionTimeInMin + dbOptions.ComputedRetentionTimeDiff),
true);
Precursor prec = new Precursor(closestTrack, spectrum.PrecursorCharge, entry);
Tracks.Add(closestTrack, prec);
Add(new Query(dbOptions, entry, spectrum, prec, NbSpectrum));
}
}//*/
if (newQueries.Count > 0)
{
//Remove precursors if estimated fragment intensities are too low (based on precursor intensity ratios and isolation window placement)
foreach (Query q in newQueries)
{
//if (q.precursor.Track.INTENSITY > intensityCumul * dbOptions.MinimumPrecursorIntensityRatioInIsolationWindow)//Need to be 5% of all intensity
//{
this.Add(q);
//}
}
}
Console.Write("\r{0}% ", ((100 * NbSpectrum) / spectra.Count));
}
Console.Write("\r{0}% ", 100);
//Sort queries to ease search
this.Sort(AscendingPrecursorMassComparison);
foreach (Track track in Tracks.Keys)
{
if (!Isotopes.ContainsKey(track))
{
Precursors.Add(Tracks[track]);
}
}
//TODO Validate this approach
//REMOVE QUERIES RELATED TO AN ISOTOPE and Compute the average CoElution
Dictionary <ProductSpectrum, double> DicOfSpectrumIntensities = new Dictionary <ProductSpectrum, double>();
for (int i = 0; i < this.Count;)
{
Query query = this[i];
if (!Isotopes.ContainsKey(query.precursor.Track))
{
if (!DicOfSpectrumIntensities.ContainsKey(query.spectrum))
{
DicOfSpectrumIntensities.Add(query.spectrum, query.precursor.Track.INTENSITY);
}
else
{
DicOfSpectrumIntensities[query.spectrum] += query.precursor.Track.INTENSITY;
}
i++;
}
else
{
this.RemoveAt(i);
}
}
//REMOVE Queries with Precursor intensities too low
for (int i = 0; i < this.Count;)
{
Query query = this[i];
if (query.precursor.Track.INTENSITY < DicOfSpectrumIntensities[query.spectrum] * dbOptions.MinimumPrecursorIntensityRatioInIsolationWindow)
{
this.RemoveAt(i);
}
else
{
i++;
}
}//*/
Dictionary <ProductSpectrum, int> DicOfSpectrumTracks = new Dictionary <ProductSpectrum, int>();
for (int i = 0; i < this.Count;)
{
Query query = this[i];
if (!Isotopes.ContainsKey(query.precursor.Track))
{
if (!DicOfSpectrumTracks.ContainsKey(query.spectrum))
{
DicOfSpectrumTracks.Add(query.spectrum, 1);
}
else
{
DicOfSpectrumTracks[query.spectrum]++;
}
i++;
}
else
{
this.RemoveAt(i);
}
}
double averageNbPrecursorPerSpectrum = 0;
int nbSpectrumMatchedToTrack = 0;
foreach (ProductSpectrum spectrum in DicOfSpectrumTracks.Keys)
{
nbSpectrumMatchedToTrack++;
averageNbPrecursorPerSpectrum += DicOfSpectrumTracks[spectrum];
}
dbOptions.ConSole.WriteLine(entry.sSDF + " :" + Precursors.Count + " precursors [" + Isotopes.Count + " isotopes] spreaded in " + Count + " queries [" + nbMissedTrack + " trackless precursors]");
dbOptions.ConSole.WriteLine("Average Precursors per Spectrum : " + averageNbPrecursorPerSpectrum / (double)nbSpectrumMatchedToTrack);
}
public static Spectra Load(pwiz.CLI.msdata.MSDataFile msFile, DBOptions options, string filePath, bool loadMS = true, bool filterMS2 = true)
{
//Find file name in msFile;
string mzMlFilepath = filePath;
int num_spectra = msFile.run.spectrumList.size();
Spectra spectra = new Spectra(num_spectra);
//List<Trail> trails = new List<Trail>();
MS1Spectrum previousMS1 = null;
try
{
//TODO DONT forget to remove the limiter
//int maxNbMSMS = 10;
double LastMs1InjectionTime = 0;
for (int i = 0; i < num_spectra /* && i < 200*/; i++)//TODO Fix that later!
{
//Spectrum
pwiz.CLI.msdata.Spectrum spec = msFile.run.spectrumList.spectrum(i, true);
if (spec.precursors.Count > 0 || spec.cvParam(pwiz.CLI.cv.CVID.MS_ms_level).value > 1)//is an MSMS
{
double retention_time = spec.scanList.scans[0].cvParam(pwiz.CLI.cv.CVID.MS_scan_start_time).timeInSeconds() / 60.0;
//List precursors and their intensities
double precursor_mz = 0;//Is there a value for the time a scan took to complete?
int charge = 2;
double precursor_intensity = 0;
string fragmentation_method = "unknown";
double isolationWindow = 1.0;
double injectionTime = spec.scanList.scans[0].cvParam(pwiz.CLI.cv.CVID.MS_ion_injection_time).value;
foreach (pwiz.CLI.msdata.Precursor precursor in spec.precursors)
{
fragmentation_method = precursor.activation.cvParams[0].name;
if (precursor.isolationWindow.cvParams.Count > 2 && (double)precursor.isolationWindow.cvParams[1].value == (double)precursor.isolationWindow.cvParams[2].value)
{
isolationWindow = precursor.isolationWindow.cvParams[1].value;
}
else if (precursor.isolationWindow.cvParams.Count > 2)
{
options.ConSole.WriteLine("Weird Isolation Window");
}
foreach (pwiz.CLI.msdata.SelectedIon ion in precursor.selectedIons)
{
//Cycle through MS to get real precursor intensities
precursor_mz = ion.cvParams[0].value;
if (ion.cvParams.Count > 1)
{
charge = (int)ion.cvParams[1].value;
}
//else
// dbOptions.ConSole.WriteLine("No charge computed for precursor ");
if (ion.cvParams.Count > 2)
{
precursor_intensity = ion.cvParams[2].value;
}
}
}
int scan_index = i;
int scan_number = scan_index + 1;
pwiz.CLI.msdata.BinaryDataArray mz = spec.getMZArray();
pwiz.CLI.msdata.BinaryDataArray intensity = spec.getIntensityArray();
int num_peaks = mz.data.Count;
if (num_peaks != intensity.data.Count)
{
options.ConSole.WriteLine("PreoteWizard reports peaks arrays (mz/intensity) of different sizes : (" + num_peaks + "/" + intensity.data.Count + ")");
if (intensity.data.Count < num_peaks)
{
num_peaks = intensity.data.Count;
}
}
GraphML_List <MsMsPeak> peaks = new GraphML_List <MsMsPeak>(num_peaks);
for (int k = 0; k < num_peaks; k++)
{
if (intensity.data[k] > 0)
{
MsMsPeak peak = new MsMsPeak(mz.data[k], intensity.data[k], 0);
peaks.Add(peak);
}
}
mz.Dispose(); mz = null;
intensity.Dispose(); intensity = null;
peaks.Sort(MsMsPeak.AscendingMzComparison);
if (filterMS2)
{
//peaks = AssignChargeStates(peaks, options.maximumAssumedPrecursorChargeState, options.precursorMassTolerance);
//peaks = Deisotopebkp(peaks, options.maximumAssumedPrecursorChargeState, options.precursorMassTolerance);
peaks = AssignChargeStatesAndDeisotope(peaks, options.MaximumPrecursorChargeState, new MassTolerance(options.productMassTolerance.Value * 0.5, options.productMassTolerance.Units));
peaks = FilterPeaks(peaks, options.MaximumNumberOfFragmentsPerSpectrum);
//TODO Add Contaminant removal
//peaks = ContaminantMasses.RemoveContaminantsFromMzSortedList(peaks, options.productMassTolerance);
//Can sometime be sorted by intensity after this call
//peaks = FilterPeaksV2(peaks);
peaks.Sort(MsMsPeak.AscendingMzComparison);
}
/*//TODO Validate that in most cases, next steps can calculate missing charge
* if (charge == 0)
* {
* for (int c = options.minimumAssumedPrecursorChargeState; c <= options.maximumAssumedPrecursorChargeState; c++)
* {
* if (options.assignChargeStates)
* {
* peaks = AssignChargeStates(peaks, c, options.productMassTolerance);
* if (options.deisotope)
* {
* peaks = Deisotope(peaks, c, options.productMassTolerance);
* }
* }
*
* double precursor_mass = Utilities.MassFromMZ(precursor_mz, c);
*
* ProductSpectrum spectrum = new ProductSpectrum(mzMlFilepath, scan_number, retention_time, fragmentation_method, precursor_mz, precursor_intensity, c, precursor_mass, peaks);
* spectra.Add(spectrum);
* }
* }
* else//*/
{/*
* if (options.assignChargeStates)
* {
* peaks = AssignChargeStatesbkp(peaks, charge, options.productMassTolerance);
* if (options.deisotope)
* {
* peaks = Deisotopebkp(peaks, charge, options.productMassTolerance);
* }
* }//*/
//peaks = AssignChargeStatesAndDeisotope(peaks, options.maximumAssumedPrecursorChargeState, options.productMassTolerance);
double precursor_mass = Numerics.MassFromMZ(precursor_mz, charge);
ProductSpectrum spectrum = new ProductSpectrum(scan_number, retention_time, fragmentation_method, precursor_mz, precursor_intensity, charge, precursor_mass, peaks, isolationWindow, injectionTime, LastMs1InjectionTime);
spectra.AddMSMS(spectrum);
//zones.Add(new Zone(precursor_mz - isolationWindow, precursor_mz + isolationWindow, retention_time));
}
//if (spectra.Count >= maxNbMSMS)
// i = 10000000;
}
else //Is an MS
{
LastMs1InjectionTime = spec.scanList.scans[0].cvParam(pwiz.CLI.cv.CVID.MS_ion_injection_time).value;
if (loadMS)
{
double retention_time = spec.scanList.scans[0].cvParam(pwiz.CLI.cv.CVID.MS_scan_start_time).timeInSeconds() / 60.0;
pwiz.CLI.msdata.BinaryDataArray mz = spec.getMZArray();
pwiz.CLI.msdata.BinaryDataArray intensity = spec.getIntensityArray();
if (previousMS1 != null)
{
previousMS1.ScanDuration = retention_time - previousMS1.RetentionTimeInMin;
spectra.MS1s.Add(previousMS1);
}
previousMS1 = new MS1Spectrum(i, retention_time, intensity.data, mz.data, 1);
//Trail.Follow(mz.data, intensity.data, retention_time, ref trails, options);
//Trail.RemoveFinished(ref trails, spectra, 1);
}
}
spec.Dispose(); spec = null;
Console.Write("\r{0}% ", ((100 * i) / num_spectra));
}
if (previousMS1 != null)
{
spectra.MS1s.Add(previousMS1);
}
/*
* //Optimization of Track following parameters
* long nbChargedTracks = 0;
* for(int missingScans = 1; missingScans < 5; missingScans++)
* {
* for(int centroid = 1; centroid < 5; centroid++)
* {
* for(int minPeaks = 1; minPeaks < 7; minPeaks++)
* {
* for(double valleyFactor = 0.1; valleyFactor < 4; valleyFactor += 0.3)
* {
* //weightedMean
* Tracks tracks = ComputeSpectraTracks(spectra, options, mzMlFilepath, missingScans, centroid, minPeaks, valleyFactor, MaxQuant.CentroidPosition.weightedMean);
* tracks.Sort(Tracks.AscendingPrecursorMassComparison);
* long cumulIsotopes = 0;
* foreach (stTrack track in tracks)
* cumulIsotopes += Queries.GetIsotopes(track, options, tracks, sample).Count;
* if (cumulIsotopes > nbChargedTracks)
* {
* nbChargedTracks = cumulIsotopes;
* dbOptions.ConSole.WriteLine(missingScans + "," + centroid + "," + minPeaks + "," + valleyFactor + ",weightedMean");
* }
*
* //Gaussian
* tracks = ComputeSpectraTracks(spectra, options, mzMlFilepath, missingScans, centroid, minPeaks, valleyFactor, MaxQuant.CentroidPosition.gaussian);
* tracks.Sort(Tracks.AscendingPrecursorMassComparison);
* cumulIsotopes = 0;
* foreach (stTrack track in tracks)
* cumulIsotopes += Queries.GetIsotopes(track, options, tracks, sample).Count;
* if (cumulIsotopes > nbChargedTracks)
* {
* nbChargedTracks = cumulIsotopes;
* dbOptions.ConSole.WriteLine(missingScans + "," + centroid + "," + minPeaks + "," + valleyFactor + ",Gaussian");
* }
* }
* }
* }
* }//*/
if (spectra.MS1s.Count > 0)
{
spectra.tracks = ComputeSpectraTracks(spectra, options, mzMlFilepath, 3, 1, 3, 1.7, MaxQuant.CentroidPosition.weightedMean);
}
else
{
spectra.tracks = new Tracks();
}
spectra.tracks.Sort(Tracks.AscendingPrecursorMassComparison);
Console.Write("\r{0}% ", 100);
//ContaminantMasses.DisplayContaminants();
}
catch (Exception ex)
{
options.ConSole.WriteLine(ex.StackTrace);
options.ConSole.WriteLine(ex.Message);
}
return(spectra);
}
public ProductSpectrum(int scanNumber, double retentionTimeInMin, string fragmentationMethod, double precursorMZ, double precursorIntensity, int precursorCharge, double precursorMass, GraphML_List <MsMsPeak> peaks, double isolationWindow, double injectionTime, double ms1InjectionTime)
{
this.INDEX = COMPTEUR++;
//this.Filename = filename;
this.ScanNumber = scanNumber;
this.RetentionTimeInMin = retentionTimeInMin;
this.FragmentationMethod = fragmentationMethod;
this.PrecursorMZ = precursorMZ;
this.PrecursorIntensity = precursorIntensity;
this.PrecursorCharge = precursorCharge;
this.PrecursorMass = precursorMass;
this.IsolationWindow = isolationWindow;
this.InjectionTime = injectionTime;
this.Ms1InjectionTime = ms1InjectionTime;
this.TotalIntensity = 0.0;
this.MostIntensePeak = 0.0;
if (peaks != null)
{
this.Peaks = peaks;
for (int p = 0; p < peaks.Count; p++)
{
if (peaks[p].Intensity > this.MostIntensePeak)
{
this.MostIntensePeak = peaks[p].Intensity;
}
this.TotalIntensity += peaks[p].Intensity;
}
}
this.PrecursorIntensityPerMilliSecond = precursorIntensity / ms1InjectionTime;
}
private static GraphML_List <MsMsPeak> AssignChargeStatesAndDeisotope(GraphML_List <MsMsPeak> peaks, int maxCharge, MassTolerance isotopicMzTolerance)
{
GraphML_List <MsMsPeak> new_peaks = new GraphML_List <MsMsPeak>(peaks);
//peaks.Sort(MSPeak.AscendingMzComparison);
int[] bestIsotopes = new int[4];
int[] currentIsotopes = new int[4];
for (int lowMassIndex = 0; lowMassIndex < new_peaks.Count - 1; lowMassIndex++)
{
double bestChargeScore = 0;
int bestCharge = 0;
bestIsotopes[0] = 0; bestIsotopes[1] = 0; bestIsotopes[2] = 0; bestIsotopes[3] = 0;
for (int charge = maxCharge; charge > 0; charge--)
{
currentIsotopes[0] = 0; currentIsotopes[1] = 0; currentIsotopes[2] = 0; currentIsotopes[3] = 0;
double score = 0;
int potentialIsotopeIndex = lowMassIndex + 1;
for (int isotope = 1; isotope <= 4; isotope++)
{
double bestMassError = isotopicMzTolerance.Value;
double aim = Numerics.IsotopicMassShift(isotope, charge) + new_peaks[lowMassIndex].MZ;
while (potentialIsotopeIndex < new_peaks.Count && new_peaks[potentialIsotopeIndex].MZ < aim + bestMassError)
{
if (new_peaks[lowMassIndex].Intensity > new_peaks[potentialIsotopeIndex].Intensity)
{
double massError = Math.Abs(Numerics.CalculateMassError(new_peaks[potentialIsotopeIndex].MZ, aim, isotopicMzTolerance.Units));
if (massError < bestMassError)
{
bestMassError = massError;
currentIsotopes[isotope - 1] = potentialIsotopeIndex;
}
}
potentialIsotopeIndex++;
}
score += isotopicMzTolerance.Value - bestMassError;
if (score == 0)
{
break;
}
;
}
if (score > bestChargeScore)
{
bestIsotopes[0] = currentIsotopes[0];
bestIsotopes[1] = currentIsotopes[1];
bestIsotopes[2] = currentIsotopes[2];
bestIsotopes[3] = currentIsotopes[3];
bestChargeScore = score;
bestCharge = charge;
}
}
new_peaks[lowMassIndex].Charge = bestCharge;
for (int i = 3; i >= 0; i--)
{
if (bestIsotopes[i] > 0)
{
new_peaks[lowMassIndex].Intensity += new_peaks[bestIsotopes[i]].Intensity;
new_peaks.RemoveAt(bestIsotopes[i]);
}
}
}
return(new_peaks);
}
public PeptideMatch()
{
clusters = new GraphML_List <Cluster>();
}
public PeptideMatch(Peptide peptide)
{
this.peptide = peptide;
clusters = new GraphML_List <Cluster>();
}
public ProductSpectrum()
{
Peaks = new GraphML_List <MsMsPeak>();
}
public ProteinGroupMatch()
: base()
{
PeptideMatches = new GraphML_List <PeptideMatch>();
Proteins = new GraphML_List <Protein>();
}