public Queries(DBOptions dbOptions)
: base()
{
MaxMZ = double.MinValue;
MinMZ = double.MaxValue;
MaxRt = double.MinValue;
MinRt = double.MaxValue;
//this.sample = sample;
this.dbOptions = dbOptions;
Precursors = new Precursors();
}
/// <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 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 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 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 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 Queries()
{
Precursors = new Precursors();
}