private bool AddEnvelope(ChromatographProfile chro, PeakList <Peak> curPeaks)
{
ChromatographProfileScan envelope = FindEnvelope(curPeaks, chro.IsotopicIons, options.MzTolerancePPM);
if (envelope.Count < options.ProfileLength)
{
return(false);
}
if (envelope.CalculateProfileCorrelation(chro.IsotopicIons) < options.MinimumCorrelation)
{
return(false);
}
chro.Profiles.Add(envelope);
return(true);
}
private bool AddEnvelope(ChromatographProfile chro, CacheRawFile rawReader, List <FullMS> fullMSList, int scanIndex)
{
PeakList <Peak> ms1 = ReadFullMS(rawReader, fullMSList, scanIndex);
ChromatographProfileScan envelope = null;
if (!profileFinder.Find(ms1, chro, options.MzTolerancePPM, options.MinimumProfileLength, ref envelope))
{
return(false);
}
if (options.MinimumProfileCorrelation > 0 && envelope.CalculateProfileCorrelation(chro.IsotopicIntensities) < options.MinimumProfileCorrelation)
{
return(false);
}
chro.Profiles.Add(envelope);
return(true);
}
protected override List <Peak> Resolve(ChromatographProfile chro, List <List <Peak> > envelope)
{
return((from peaks in envelope
let peakindex = peaks.FindMaxIndex()
select peaks[peakindex]).ToList());
}
protected override List <Peak> Resolve(ChromatographProfile chro, List <List <Peak> > envelope)
{
//return Isotopic.InterfernceOptimization.ResolveByPearsonCorrelation(chro.IsotopicIntensities, envelope);
return(Isotopic.InterfernceOptimization.ResolveByKullbackLeiblerDistance(chro.IsotopicIntensities, envelope));
}
public override IEnumerable <string> Process()
{
var format = new MascotPeptideTextFormat();
var spectra = format.ReadFromFile(options.InputFile);
var peptideMap = spectra.ToGroupDictionary(m => m.Query.FileScan.Experimental.ToLower());
var rawfiles = Directory.GetFiles(options.RawDirectory, "*.raw", SearchOption.AllDirectories).ToDictionary(m => Path.GetFileNameWithoutExtension(m).ToLower());
var rententionWindow = options.RetentionTimeWindow;
var missed = peptideMap.Keys.Except(rawfiles.Keys).ToArray();
if (missed.Length > 0)
{
throw new Exception(string.Format("Cannot find raw file of {0} in directory {1}", missed.Merge("/"), options.RawDirectory));
}
var option = new ParallelOptions()
{
//MaxDegreeOfParallelism = Math.Min(1, peptideMap.Count),
MaxDegreeOfParallelism = Math.Min(Environment.ProcessorCount, peptideMap.Count),
};
Parallel.ForEach(peptideMap, option, raw =>
{
//foreach (var raw in peptideMap)
//{
var peptides = raw.Value;
Progress.SetMessage("Preparing isotopic for " + raw.Key + " ...");
var waitingPeaks = new List <ChromatographProfile>();
foreach (var peptide in peptides)
{
string file = GetTargetFile(peptide);
var chro = new ChromatographProfile()
{
Experimental = peptide.Query.FileScan.Experimental,
IdentifiedScan = peptide.Query.FileScan.FirstScan,
ObservedMz = peptide.GetPrecursorMz(),
TheoreticalMz = peptide.GetTheoreticalMz(),
Charge = peptide.Query.Charge,
Sequence = peptide.Peptide.PureSequence,
FileName = Path.GetFileName(file)
};
chro.InitializeIsotopicIons(options.MzTolerancePPM);
waitingPeaks.Add(chro);
}
if (waitingPeaks.Count == 0)
{
//continue;
return;
}
Dictionary <string, List <ChromatographProfile> > resultMap = new Dictionary <string, List <ChromatographProfile> >();
List <FullMS> fullMSList = new List <FullMS>();
Progress.SetMessage("Reading full ms list from " + rawfiles[raw.Key] + "...");
using (var rawReader = new CacheRawFile(RawFileFactory.GetRawFileReader(rawfiles[raw.Key])))
{
var firstScan = rawReader.GetFirstSpectrumNumber();
var lastScan = rawReader.GetLastSpectrumNumber();
for (int scan = firstScan; scan <= lastScan; scan++)
{
var mslevel = rawReader.GetMsLevel(scan);
if (mslevel == 1)
{
fullMSList.Add(new FullMS()
{
Scan = scan,
RetentionTime = rawReader.ScanToRetentionTime(scan),
Peaks = null
});
}
}
var chroGroups = waitingPeaks.GroupBy(chro => string.Format("{0}_{1:0.0000}", chro.Sequence, chro.TheoreticalMz));
foreach (var chroGroup in chroGroups)
{
List <ChromatographProfile> profileChros = new List <ChromatographProfile>();
foreach (var chro in chroGroup.OrderBy(m => m.IdentifiedScan))
{
var masterScanIndex = 0;
for (int i = 1; i < fullMSList.Count; i++)
{
if (chro.IdentifiedScan < fullMSList[i].Scan)
{
break;
}
masterScanIndex = i;
}
var masterScan = fullMSList[masterScanIndex].Scan;
var masterRetentionTime = fullMSList[masterScanIndex].RetentionTime;
bool bExist = false;
foreach (var profileChro in profileChros)
{
foreach (var pkl in profileChro.Profiles)
{
if (pkl.Scan == fullMSList[masterScanIndex].Scan)
{
pkl.Identified = true;
bExist = true;
break;
}
}
if (bExist)
{
break;
}
}
if (bExist)
{
continue;
}
Progress.SetMessage("Processing {0} : {1:0.#####} : {2} : {3}", chro.Sequence, chro.ObservedMz, chro.IdentifiedScan, Path.GetFileName(chro.FileName));
for (int scanIndex = masterScanIndex; scanIndex >= 0; scanIndex--)
{
if (Progress.IsCancellationPending())
{
throw new UserTerminatedException();
}
var curRetentionTime = fullMSList[scanIndex].RetentionTime;
if (masterRetentionTime - curRetentionTime > rententionWindow)
{
break;
}
if (!AddEnvelope(chro, rawReader, fullMSList, scanIndex))
{
break;
}
if (scanIndex == masterScanIndex)
{
chro.Profiles.Last().Identified = true;
}
}
chro.Profiles.Reverse();
for (int scanIndex = masterScanIndex + 1; scanIndex < fullMSList.Count; scanIndex++)
{
if (Progress.IsCancellationPending())
{
throw new UserTerminatedException();
}
var curRetentionTime = fullMSList[scanIndex].RetentionTime;
if (curRetentionTime - masterRetentionTime > rententionWindow)
{
break;
}
if (!AddEnvelope(chro, rawReader, fullMSList, scanIndex))
{
break;
}
}
profileChros.Add(chro);
}
profileChros.RemoveAll(l => l.Profiles.Count < options.MinimumScanCount);
profileChros.Sort((m1, m2) => m2.Profiles.Count.CompareTo(m1.Profiles.Count));
bool bMain = true;
foreach (var chro in profileChros)
{
string filename;
if (bMain)
{
filename = Path.Combine(GetTargetDirectory(chro.Experimental), chro.FileName);
}
else
{
filename = Path.Combine(GetTargetSubDirectory(chro.Experimental), Path.ChangeExtension(chro.FileName, ".sub" + Path.GetExtension(chro.FileName)));
}
bMain = false;
new ChromatographProfileTextWriter().WriteToFile(filename, chro);
new ChromatographProfileXmlFormat().WriteToFile(filename + ".xml", chro);
}
}
}
}
);
Progress.SetMessage("Finding boundaries ...");
var boundaryOptions = new RTemplateProcessorOptions()
{
InputFile = targetDir,
OutputFile = options.OutputFile,
RTemplate = BoundaryR,
RExecute = SystemUtils.GetRExecuteLocation(),
CreateNoWindow = true
};
new RTemplateProcessor(boundaryOptions)
{
Progress = this.Progress
}.Process();
return(new string[] { options.OutputFile });
}
private bool AddEnvelope(ChromatographProfile chro, CacheRawFile rawReader, List <FullMS> fullMSList, int scanIndex)
{
PeakList <Peak> fullMS = ReadFullMS(rawReader, fullMSList, scanIndex);
return(AddEnvelope(chro, fullMS));
}
public override IEnumerable <string> Process()
{
var boundaryInput = Path.ChangeExtension(options.OutputFile, ".chros.tsv");
if (!File.Exists(boundaryInput) || options.Overwrite)
{
var format = GetPeptideReader();
var spectra = format.ReadFromFile(options.InputFile);
var peptideMap = spectra.ToGroupDictionary(m => m.Query.FileScan.Experimental.ToLower());
var rawfiles = options.RawFiles.ToDictionary(m => RawFileFactory.GetExperimental(m).ToLower());
var rententionWindow = options.MaximumRetentionTimeWindow;
var missed = peptideMap.Keys.Except(rawfiles.Keys).ToArray();
if (missed.Length > 0)
{
throw new Exception(string.Format("Cannot find raw file of {0} in file list", missed.Merge("/")));
}
var optionThreadCount = options.ThreadCount == 0 ? Environment.ProcessorCount : options.ThreadCount;
var option = new ParallelOptions()
{
MaxDegreeOfParallelism = Math.Min(optionThreadCount, peptideMap.Count),
};
var chroMap = new List <Tuple <string, List <ChromatographProfile> > >();
foreach (var raw in peptideMap)
{
var peptides = raw.Value;
var waitingPeaks = new List <ChromatographProfile>();
foreach (var peptide in peptides)
{
var chro = new ChromatographProfile()
{
Experimental = peptide.Query.FileScan.Experimental,
IdentifiedScan = peptide.Query.FileScan.FirstScan,
IdentifiedRetentionTime = peptide.Query.FileScan.RetentionTime,
ObservedMz = peptide.GetPrecursorMz(),
TheoreticalMz = peptide.GetTheoreticalMz(),
Charge = peptide.Query.Charge,
Sequence = peptide.Peptide.PureSequence,
FileName = GetTargetFile(peptide),
SubFileName = GetTargetSubFile(peptide)
};
chro.InitializeIsotopicIons(options.MzTolerancePPM, options.MinimumIsotopicPercentage);
waitingPeaks.Add(chro);
}
chroMap.Add(new Tuple <string, List <ChromatographProfile> >(raw.Key, waitingPeaks));
}
ConcurrentBag <ChromatographProfile> detected = new ConcurrentBag <ChromatographProfile>();
Parallel.ForEach(chroMap, option, raw =>
{
var rawFileName = raw.Item1;
var waitingPeaks = raw.Item2;
Dictionary <string, List <ChromatographProfile> > resultMap = new Dictionary <string, List <ChromatographProfile> >();
List <FullMS> fullMSList = new List <FullMS>();
Progress.SetMessage("Reading full ms list from " + rawfiles[rawFileName] + "...");
using (var rawReader = new CacheRawFile(RawFileFactory.GetRawFileReader(rawfiles[rawFileName])))
{
var firstScan = rawReader.GetFirstSpectrumNumber();
var lastScan = rawReader.GetLastSpectrumNumber();
for (int scan = firstScan; scan <= lastScan; scan++)
{
var mslevel = rawReader.GetMsLevel(scan);
if (mslevel == 1)
{
fullMSList.Add(new FullMS()
{
Scan = scan,
RetentionTime = rawReader.ScanToRetentionTime(scan),
Peaks = null
});
}
}
foreach (var chro in waitingPeaks)
{
if (chro.IdentifiedScan == 0 && chro.IdentifiedRetentionTime > 0)
{
for (int i = 1; i < fullMSList.Count; i++)
{
if (chro.IdentifiedRetentionTime < fullMSList[i].RetentionTime)
{
break;
}
chro.IdentifiedScan = fullMSList[i].Scan + 1;
}
}
}
var chroGroups = waitingPeaks.GroupBy(chro => chro.GetPeptideId());
foreach (var chroGroup in chroGroups)
{
List <ChromatographProfile> profileChros = new List <ChromatographProfile>();
foreach (var chro in chroGroup.OrderBy(m => m.IdentifiedScan))
{
var masterScanIndex = 0;
for (int i = 1; i < fullMSList.Count; i++)
{
if (chro.IdentifiedScan < fullMSList[i].Scan)
{
break;
}
masterScanIndex = i;
}
var masterScan = fullMSList[masterScanIndex].Scan;
var masterRetentionTime = fullMSList[masterScanIndex].RetentionTime;
bool bExist = false;
foreach (var profileChro in profileChros)
{
foreach (var pkl in profileChro.Profiles)
{
if (pkl.Scan == fullMSList[masterScanIndex].Scan)
{
pkl.Identified = true;
bExist = true;
break;
}
}
if (bExist)
{
break;
}
}
if (bExist)
{
continue;
}
//Progress.SetMessage("Processing {0} : {1:0.#####} : {2} : {3}", chro.Sequence, chro.ObservedMz, chro.IdentifiedScan, Path.GetFileName(chro.FileName));
//allow one missed scan
int naCount = 2;
for (int scanIndex = masterScanIndex; scanIndex >= 0; scanIndex--)
{
if (Progress.IsCancellationPending())
{
throw new UserTerminatedException();
}
var curRetentionTime = fullMSList[scanIndex].RetentionTime;
if (masterRetentionTime - curRetentionTime > rententionWindow)
{
break;
}
if (!AddEnvelope(chro, rawReader, fullMSList, scanIndex))
{
naCount--;
if (naCount == 0)
{
break;
}
else
{
continue;
}
}
if (scanIndex == masterScanIndex)
{
chro.Profiles.Last().Identified = true;
}
}
chro.Profiles.Reverse();
naCount = 2;
for (int scanIndex = masterScanIndex + 1; scanIndex < fullMSList.Count; scanIndex++)
{
if (Progress.IsCancellationPending())
{
throw new UserTerminatedException();
}
var curRetentionTime = fullMSList[scanIndex].RetentionTime;
if (curRetentionTime - masterRetentionTime > rententionWindow)
{
break;
}
if (!AddEnvelope(chro, rawReader, fullMSList, scanIndex))
{
naCount--;
if (naCount == 0)
{
break;
}
else
{
continue;
}
}
}
profileChros.Add(chro);
}
profileChros.RemoveAll(l => l.Profiles.Count < options.MinimumScanCount);
profileChros.Sort((m1, m2) => m2.Profiles.Count.CompareTo(m1.Profiles.Count));
bool bMain = true;
foreach (var chro in profileChros)
{
var filename = bMain ? chro.FileName : chro.SubFileName;
if (bMain)
{
detected.Add(chro);
}
bMain = false;
new ChromatographProfileTextWriter().WriteToFile(filename, chro);
new ChromatographProfileXmlFormat().WriteToFile(filename + ".xml", chro);
}
}
}
}
);
var chroList = new List <ChromatographProfile>(detected);
chroList.Sort((m1, m2) => m1.FileName.CompareTo(m2.FileName));
if (chroList.Count == 0)
{
throw new Exception("Cannot find chromotograph!");
}
using (var sw = new StreamWriter(boundaryInput))
{
sw.WriteLine("ChroDirectory\tChroFile\tSample\tPeptideId\tTheoreticalMz\tCharge\tIdentifiedScan");
foreach (var chro in chroList)
{
sw.WriteLine("{0}\t{1}\t{2}\t{3}\t{4}\t{5}\t{6}",
Path.GetDirectoryName(chro.FileName).Replace("\\", "/"),
Path.GetFileNameWithoutExtension(chro.FileName),
chro.Experimental,
chro.GetPeptideId(),
chro.TheoreticalMz,
chro.Charge,
chro.IdentifiedScan);
}
}
}
if (!File.Exists(options.OutputFile) || options.Overwrite)
{
Progress.SetMessage("Finding boundaries ...");
var boundaryOptions = new RTemplateProcessorOptions()
{
InputFile = boundaryInput,
OutputFile = options.OutputFile,
RTemplate = BoundaryR,
RExecute = ExternalProgramConfig.GetExternalProgram("R"),
CreateNoWindow = true
};
boundaryOptions.Parameters.Add("outputImage<-" + (options.DrawImage ? "1" : "0"));
boundaryOptions.Parameters.Add("maximumProfileDistance<-" + options.MaximumProfileDistance.ToString());
new RTemplateProcessor(boundaryOptions)
{
Progress = this.Progress
}.Process();
}
//if (options.DrawImage)
//{
// Progress.SetMessage("Drawing images ...");
// var imageOptions = new RTemplateProcessorOptions()
// {
// InputFile = options.OutputFile,
// OutputFile = Path.ChangeExtension(options.OutputFile, ".image"),
// RTemplate = ImageR,
// RExecute = SystemUtils.GetRExecuteLocation(),
// CreateNoWindow = true,
// NoResultFile = true
// };
// new RTemplateProcessor(imageOptions) { Progress = this.Progress }.Process();
//}
return(new string[] { options.OutputFile });
}
protected abstract List <Peak> Resolve(ChromatographProfile chro, List <List <Peak> > envelope);
public bool Find(PeakList <Peak> ms1, ChromatographProfile chro, double mzTolerancePPM, int minimumProfileLength, ref ChromatographProfileScan result)
{
var rawPeaks = ms1.GetRange(chro.IsotopicIons[0].MinimumMzWithinTolerance, chro.IsotopicIons[chro.IsotopicIons.Length - 1].MaximumMzWithinTolerance);
if (rawPeaks.Count < minimumProfileLength)
{
return(false);
}
List <List <Peak> > envelope = new List <List <Peak> >();
int peakIndex = 0;
foreach (var peak in chro.IsotopicIons)
{
List <Peak> findPeaks = new List <Peak>();
while (peakIndex < rawPeaks.Count)
{
var curPeak = rawPeaks[peakIndex];
if (curPeak.Mz < peak.MinimumMzWithinTolerance)
{
peakIndex++;
continue;
}
if (curPeak.Mz > peak.MaximumMzWithinTolerance)
{
break;
}
findPeaks.Add(curPeak);
peakIndex++;
}
if (findPeaks.Count == 0)
{
break;
}
else
{
envelope.Add(findPeaks);
}
}
if (envelope.Count < minimumProfileLength)
{
return(false);
}
result = new ChromatographProfileScan();
result.Scan = ms1.FirstScan;
result.RetentionTime = ms1.ScanTimes[0].RetentionTime;
result.RawPeaks = new List <Peak>();
List <Peak> resolved = Resolve(chro, envelope);
foreach (var findPeak in rawPeaks)
{
var isotopicIndex = resolved.IndexOf(findPeak);
if (isotopicIndex == -1)
{
result.RawPeaks.Add(findPeak);
}
else
{
var ion = chro.IsotopicIons[isotopicIndex];
result.Add(new ChromatographProfileScanPeak()
{
Isotopic = isotopicIndex + 1,
Mz = findPeak.Mz,
Intensity = findPeak.Intensity,
Noise = findPeak.Noise,
PPMDistance = PrecursorUtils.mz2ppm(findPeak.Mz, findPeak.Mz - ion.Mz)
});
}
}
return(true);
}
