/// <summary>
/// Constructor used for Unit tests
/// </summary>
/// <param name="masses"></param>
public Queries(DBOptions options, double[] masses)
{
Precursors = new Precursors();
foreach (double mass in masses)
{
Precursor precursor = new Precursor();
precursor.Mass = mass;
Add(new Query(options, null, null, precursor));
Precursors.Add(precursor);
}
}
public void CumulPsm(List <Precursor> matches)
{
foreach (Precursor precursor in matches)
{
Sample entry = precursor.sample;
precursors.Add(precursor);//Remove blank psm (unmatched spectrum/query duos)
/* if (!samples.ContainsKey(entry.PROJECT.CONDITION))
* samples.Add(entry.PROJECT.CONDITION, new List<int>());
* if (!samples[entry.PROJECT.CONDITION].Contains(entry.PROJECT.REPLICATE))
* samples[entry.PROJECT.CONDITION].Add(entry.PROJECT.REPLICATE);//*/
}
}
public long SetPrecursors(Precursors precursors)
{
this.precursors = precursors;
this.matchedPrecursors = new Precursors();
long nbTargets = 0;
foreach (Precursor precursor in precursors)
{
if (precursor.psms != null && precursor.psms.Count > 0)
{
if (precursor.Target)
{
nbTargets++;
}
matchedPrecursors.Add(precursor);
}
}
return(nbTargets);
}
public void Export(double fdr, string keyword = "", bool onlyPrecursors = false)
{
dbOptions.ConSole.WriteLine("Exporting at " + (fdr * 100) + "% FDR (Decoy/Target)...");
List <Precursor> prec = null;
if (precursors != null)
{
if (matchedPrecursors == null)
{
this.matchedPrecursors = new Precursors();
foreach (Precursor precursor in precursors)
{
if (precursor.psms.Count > 0)
{
matchedPrecursors.Add(precursor);
}
}
}/*
* List<Precursor> prec = FDR.PrecursorsV2(precursors, fdr, 1);
* Sol.CONSOLE.OutputLine("> " + prec.Count + " Precursors");
* MSSearcher.Export(dbOptions.outputFolder + keyword + "precursors.csv", prec);//*/
/*
* prec = Optimizer.PrecursorOptimizer(matchedPrecursors, fdr);
* Sol.CONSOLE.OutputLine("> " + prec.Count + " Optimized Precursors");
* MSSearcher.Export(dbOptions.outputFolder + keyword + "Optimized_precursors.csv", prec);//*/
prec = matchedPrecursors.ComputeAtFDR(fdr);
dbOptions.ConSole.WriteLine("> " + prec.Count + " Uptimized V5 Precursors");
MSSearcher.Export(dbOptions.OutputFolder + keyword + "UptimizedV5_precursors.csv", prec);
}
if (!onlyPrecursors)
{
if (queries != null)
{
List <Query> qs = queries.ComputeAtFDR(fdr);
dbOptions.ConSole.WriteLine("> " + qs.Count + " PSMs (Top 10)");
MSSearcher.Export(dbOptions.OutputFolder + keyword + "queries.csv", qs);
}
if (clusters != null)
{
//List<PeptideSpectrumMatch> psms = FDR.PSMs(clusters, fdr, 1, 10);
//dbOptions.ConSole.WriteLine("> " + psms.Count + " PSMs (Top 10)");
//MSSearcher.Export(dbOptions.outputFolder + keyword + "psms_Top10.csv", psms);
//psms = FDR.PSMs(clusters, fdr, 1, 1);
//dbOptions.ConSole.WriteLine("> " + psms.Count + " PSMs");
//MSSearcher.Export(dbOptions.outputFolder + keyword + "psms_Best.csv", psms);
}
if (peptides != null)
{
List <PeptideMatch> pep = peptideSequences.ComputeAtFDR(fdr);
dbOptions.ConSole.WriteLine("> " + pep.Count + " Peptides Sequences (Version 5)");
PeptideSearcher.Export(dbOptions.OutputFolder + keyword + "peptideSequencesV5_.csv", pep);
PeptideSearcher sr = new PeptideSearcher(dbOptions);
PeptideMatches seqs = sr.Search(clusters, prec, false);
pep = seqs.ComputeAtFDR(fdr);
dbOptions.ConSole.WriteLine("> " + pep.Count + " Peptides Sequences (Version 5b)");
PeptideSearcher.Export(dbOptions.OutputFolder + keyword + "peptideSequencesV5b_PrecursorFDRed.csv", pep);
pep = peptides.ComputeAtFDR(fdr);
dbOptions.ConSole.WriteLine("> " + pep.Count + " Peptides (Version 5)");
PeptideSearcher.Export(dbOptions.OutputFolder + keyword + "peptidesV5_.csv", pep);
}
if (proteins != null)
{
List <ProteinGroupMatch> prots = proteins.ComputeAtFDR(fdr);
dbOptions.ConSole.WriteLine("> " + prots.Count + " Proteins");
ProteinSearcher.Export(dbOptions.OutputFolder + keyword + "proteins_.csv", prots);
}
}
}
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);
}
