public IEnumerable <Tuple <Peptide, int> > DigestProteomeOnTheFlyFast(List <Protein> proteins, bool allowSNP, Queries AllQueries)
{
foreach (Peptide peptide in ProteinSearcher.ProteinDigest(options, proteins, allowSNP))
{
//int firstIndex = AllQueries.BinarySearch(MassTolerance.MzFloor(peptide.MonoisotopicMass, options.precursorMassTolerance));
//if (firstIndex >= 0 && firstIndex < AllQueries.Count)
// yield return new Tuple<Peptide, int>(peptide, firstIndex);
//foreach (Peptide peptide in ProteinSearcher.ProteinDigestNoEnzyme(dbOptions, proteins, AllQueries))
//if (!TargetPeptides.Contains(peptide.BaseSequence))
//{
foreach (Peptide modPeptide in peptide.GetVariablyModifiedPeptides(options.variableModifications, options.maximumVariableModificationIsoforms))
{
modPeptide.SetFixedModifications(options.fixedModifications);
int firstIndex = AllQueries.BinarySearch(MassTolerance.MzFloor(modPeptide.MonoisotopicMass, options.precursorMassTolerance));
if (firstIndex >= 0 && firstIndex < AllQueries.Count)
{
yield return(new Tuple <Peptide, int>(modPeptide, firstIndex));
}
}
//TODO check if this favors targets over decoys since proteins are sorted target->deco
// TargetPeptides.Add(peptide.BaseSequence);
//}
}
}
public IEnumerable <Tuple <Peptide, int> > DigestProteomeOnTheFly(List <Protein> proteins, bool allowSNP, Queries AllQueries)
{
//HashSet<string> TargetPeptides = new HashSet<string>();
//Dictionary<string, int> TargetPeptides = new Dictionary<string, int>();
//Digest proteins and store peptides in a Dictionnary
//Does not fit in memory -> 360 Go ....
//dicOfPeptideSequences = new Dictionary<string, List<Protein>>();
//double minimumMonoisotopicPeakOffset = dbOptions.precursorMonoisotopicPeakCorrection ? dbOptions.minimumPrecursorMonoisotopicPeakOffset : 0;
//double maximumMonoisotopicPeakOffset = dbOptions.precursorMonoisotopicPeakCorrection ? dbOptions.maximumPrecursorMonoisotopicPeakOffset : 0;
foreach (Peptide peptide in ProteinSearcher.ProteinDigest(options, proteins, allowSNP))
{
//int firstIndex = AllQueries.BinarySearch(MassTolerance.MzFloor(peptide.MonoisotopicMass, options.precursorMassTolerance));
//if (firstIndex >= 0 && firstIndex < AllQueries.Count)
// yield return new Tuple<Peptide, int>(peptide, firstIndex);
//foreach (Peptide peptide in ProteinSearcher.ProteinDigestNoEnzyme(dbOptions, proteins, AllQueries))
//if (!TargetPeptides.Contains(peptide.BaseSequence))
//{
foreach (Peptide modPeptide in peptide.GetVariablyModifiedPeptides(options.variableModifications, options.maximumVariableModificationIsoforms))
{
modPeptide.SetFixedModifications(options.fixedModifications);
int firstIndex = AllQueries.BinarySearch(MassTolerance.MzFloor(modPeptide.MonoisotopicMass, options.precursorMassTolerance));
if (firstIndex >= 0 && firstIndex < AllQueries.Count)
{
yield return(new Tuple <Peptide, int>(modPeptide, firstIndex));
}
}
//TODO check if this favors targets over decoys since proteins are sorted target->deco
// TargetPeptides.Add(peptide.BaseSequence);
//}
}
}
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);
}
}
}
}//*/
/// <summary>
/// Latest version of the search routine. Associates spectrum to digested peptide sequences,
/// aligns precursors and fragments, clusters common precursors accross samples, create the list of detected
/// peptide and protein sequences.
/// TODO Align maps together (retention time)
/// </summary>
/// <param name="queries"></param>
/// <returns></returns>
public Result SearchVersionAugust2013(Queries queries, bool optimize)
{
Result result = new Result();
result.queries = queries;
result.dbOptions = dbOptions;
result.samples = Project;
DBSearcher dbSearcher = new DBSearcher(dbOptions);
Digestion ps = new Digestion(dbOptions);
if (dbOptions.NoEnzymeSearch)
{
result.SetPrecursors(dbSearcher.Search(queries, ps.DigestProteomeOnTheFlyNoEnzyme(AllProteins, queries)));
}
else
{
result.SetPrecursors(dbSearcher.Search(queries, ps.DigestProteomeOnTheFly(AllProteins, false, queries)));
}
dbOptions.ConSole.WriteLine(result.precursors.Count + " precursors matched !");
PeptideSpectrumMatches allPSMs = new PeptideSpectrumMatches();
foreach (Precursor precursor in result.precursors)
{
foreach (PeptideSpectrumMatch psm in precursor.psms_AllPossibilities)
{
allPSMs.Add(psm);
}
}
//Add all psm possibilities to psms list
foreach (PeptideSpectrumMatch psm in allPSMs)
{
psm.Query.precursor.psms.Add(psm);
}
long nbTargets = result.SetPrecursors(result.precursors);
dbOptions.ConSole.WriteLine("Targets before Optimizing Score Ratios : " + nbTargets + " [" + result.matchedPrecursors.Count + "]");
if (optimize)
{
//allPSMs.OptimizePSMScoreRatios(dbOptions, dbOptions.PSMFalseDiscoveryRate, result);
//result.matchedPrecursors.OptimizePSMScoreRatios(dbOptions, dbOptions.PSMFalseDiscoveryRate, result);
nbTargets = result.SetPrecursors(result.precursors);
dbOptions.ConSole.WriteLine("Targets : " + nbTargets + " [" + result.matchedPrecursors.Count + "]");
//*/
//TODO Improve alignment results
/*
* Align.AlignPrecursorsByDiff(result, allPSMs);
* nbTargets = result.SetPrecursors(result.precursors);
* dbOptions.ConSole.WriteLine("Targets after precursor alignment : " + nbTargets + " [" + result.matchedPrecursors.Count + "]");
*
* Align.AlignProductsByDiff(result, allPSMs);
* nbTargets = result.SetPrecursors(result.precursors);
* dbOptions.ConSole.WriteLine("Targets after fragment alignment : " + nbTargets + " [" + result.matchedPrecursors.Count + "]");
* //*/
/*
* dbOptions.precursorMassTolerance.Value = Align.CropPrecursors(result, allPSMs);
* nbTargets = result.SetPrecursors(result.precursors);
* dbOptions.ConSole.WriteLine("Targets after croping precursors : " + nbTargets + " [" + result.matchedPrecursors.Count + "]");
*
* dbOptions.productMassTolerance.Value = Align.CropProducts(result, allPSMs);
* nbTargets = result.SetPrecursors(result.precursors);
* dbOptions.ConSole.WriteLine("Targets after croping fragments : " + nbTargets + " [" + result.matchedPrecursors.Count + "]");
* //*/
}
//*/
allPSMs = null;
long bestTargets = nbTargets;
MSSearcher msSearcher = new MSSearcher(dbOptions, Project);
msSearcher.CumulPsm(result.matchedPrecursors);//TODO Check if its still needed
//Step 1 : Cluster psms together based on precursor feature //TODO Implement ProteoProfile Scoring based clustering
//Group in clusters
result.clusters = msSearcher.Search(result.matchedPrecursors, true);
//Todo Align retention times
//Todo redo clusterization, based on retention time aligned maps
//Step 2 : Regroup based on peptide sequence (Morpheus code)
PeptideSearcher pepSearcher = new PeptideSearcher(dbOptions);
result.peptides = pepSearcher.Search(result.clusters, result.matchedPrecursors, true);
result.peptideSequences = pepSearcher.Search(result.clusters, result.matchedPrecursors, false);
//Step 3 : Regroup based on protein sequences (Morpheus code)
ProteinSearcher protSearcher = new ProteinSearcher(dbOptions);
result.proteins = protSearcher.SearchLatest(result.peptideSequences);//, dbSearcher.DicOfProteins);
//long lastNbTarget = nbTargets;
//do
//{
// lastNbTarget = nbTargets;
//UpdatePsmScores(result.proteins);
// nbTargets = result.SetPrecursors(result.precursors);
//} while (lastNbTarget < nbTargets);//*/
result.peptides = pepSearcher.Search(result.clusters, result.matchedPrecursors, true);
result.peptideSequences = pepSearcher.Search(result.clusters, result.matchedPrecursors, false);
if (optimize)
{
dbOptions.ConSole.WriteLine("Targets before second Optimization of Score Ratios : " + nbTargets + " [" + result.matchedPrecursors.Count + "]");
//result.matchedPrecursors.OptimizePSMScoreRatios(dbOptions, dbOptions.PSMFalseDiscoveryRate, result);
nbTargets = result.SetPrecursors(result.precursors);
dbOptions.ConSole.WriteLine("Targets after ReOptimizing PSM Score Ratios : " + nbTargets + " [" + result.matchedPrecursors.Count + "]");//*/
}
//Step 5 : Compute the new number of Targets
nbTargets = result.SetPrecursors(result.precursors);
if (nbTargets < bestTargets)
{
dbOptions.ConSole.WriteLine("FAILED to improve PSMs while adding protein and peptide information");
}
dbOptions.ConSole.WriteLine(result.matchedPrecursors.Count + " precursors remaining after ProPheus Search!");
return(result);
}//*/