/// <summary>
/// Links a list of peptides to the features provided if the dataset has knowledge of the sequence file file
/// </summary>
private void LinkPeptidesToFeatures(string sequencePath, IEnumerable<UMCLight> aligneeFeatures, double fdr,
double idScore)
{
// Get the peptides associated with this feature set.
var peptideReaderY = PeptideReaderFactory.CreateReader(sequencePath);
if (peptideReaderY == null)
return;
// Load the peptide Y
UpdateStatus("Linking peptides to ms/ms");
var linker = new PeptideMsMsLinker();
var peptides = peptideReaderY.Read(sequencePath).ToList();
var filteredPeptides = peptides.ToList().Where(x => PeptideUtility.PassesCutoff(x, idScore, fdr)).ToList();
var msnSpectra = new List<MSSpectra>();
foreach (var feature in aligneeFeatures)
{
foreach (var msFeature in feature.MsFeatures)
{
msnSpectra.AddRange(msFeature.MSnSpectra);
}
}
linker.LinkPeptidesToSpectra(msnSpectra, filteredPeptides);
}
/// <summary>
/// Load a single dataset from the provider.
/// </summary>
/// <returns></returns>
public IList<UMCLight> LoadDataset(DatasetInformation dataset,
MsFeatureFilteringOptions msFilteringOptions,
LcmsFeatureFindingOptions lcmsFindingOptions,
LcmsFeatureFilteringOptions lcmsFilteringOptions)
{
UpdateStatus(string.Format("[{0}] - Loading dataset [{0}] - {1}.", dataset.DatasetId, dataset.DatasetName));
var datasetId = dataset.DatasetId;
var features = UmcLoaderFactory.LoadUmcFeatureData(dataset.Features.Path, dataset.DatasetId,
Providers.FeatureCache);
UpdateStatus(string.Format("[{0}] Loading MS Feature Data [{0}] - {1}.", dataset.DatasetId,
dataset.DatasetName));
var msFeatures = UmcLoaderFactory.LoadMsFeatureData(dataset.Features.Path);
var scansInfo = UmcLoaderFactory.LoadScanSummaries(dataset.Scans.Path);
dataset.BuildScanTimes(scansInfo);
var msnSpectra = new List<MSSpectra>();
// If we don't have any features, then we have to create some from the MS features
// provided to us.
if (features.Count < 1)
{
msFeatures = LcmsFeatureFilters.FilterMsFeatures(msFeatures, msFilteringOptions);
msFeatures = Filter(msFeatures, ref dataset);
features = CreateLcmsFeatures(dataset,
msFeatures,
lcmsFindingOptions,
lcmsFilteringOptions);
//var maxScan = Convert.ToDouble(features.Max(feature => feature.Scan));
//var minScan = Convert.ToDouble(features.Min(feature => feature.Scan));
var maxScan = features.Max(feature => feature.Scan);
var minScan = features.Min(feature => feature.Scan);
var id = 0;
var scanTimes = dataset.ScanTimes;
foreach (var feature in features)
{
feature.Id = id++;
//feature.Net = (Convert.ToDouble(feature.Scan) - minScan) / (maxScan - minScan);
feature.Net = (Convert.ToDouble(scanTimes[feature.Scan]) - scanTimes[minScan]) / (scanTimes[maxScan] - scanTimes[minScan]);
feature.MassMonoisotopicAligned = feature.MassMonoisotopic;
feature.NetAligned = feature.Net;
feature.GroupId = datasetId;
feature.SpectralCount = feature.MsFeatures.Count;
foreach (var msFeature in feature.MsFeatures.Where(msFeature => msFeature != null))
{
msFeature.UmcId = feature.Id;
msFeature.GroupId = datasetId;
msFeature.MSnSpectra.ForEach(x => x.GroupId = datasetId);
msnSpectra.AddRange(msFeature.MSnSpectra);
}
}
}
else
{
if (!UmcLoaderFactory.AreExistingFeatures(dataset.Features.Path))
{
var i = 0;
foreach (var feature in features)
{
feature.GroupId = datasetId;
feature.Id = i++;
}
}
// Otherwise, we need to map the MS features to the LCMS Features provided.
// This would mean that we extracted data from an existing database.
if (msFeatures.Count > 0)
{
var map = FeatureDataConverters.MapFeature(features);
foreach (var feature in
from feature in msFeatures
let doesFeatureExists = map.ContainsKey(feature.UmcId)
where doesFeatureExists
select feature)
{
map[feature.UmcId].AddChildFeature(feature);
}
}
}
// Process the MS/MS data with peptides
UpdateStatus("Reading List of Peptides");
var sequenceProvider = PeptideReaderFactory.CreateReader(dataset.SequencePath);
if (sequenceProvider != null)
{
UpdateStatus("Reading List of Peptides");
var peptides = sequenceProvider.Read(dataset.SequencePath);
var count = 0;
var peptideList = peptides.ToList();
peptideList.ForEach(x => x.Id = count++);
UpdateStatus("Linking MS/MS to any known Peptide/Metabolite Sequences");
var linker = new PeptideMsMsLinker();
linker.LinkPeptidesToSpectra(msnSpectra, peptideList);
}
return features;
}
