private PeptideGroupDocNode Refine(PeptideGroupDocNode nodePepGroup,
SrmDocument document,
ICollection<int> outlierIds,
ICollection<RefinementIdentity> includedPeptides,
ICollection<RefinementIdentity> repeatedPeptides,
Dictionary<RefinementIdentity, List<int>> acceptedPeptides,
SrmSettingsChangeMonitor progressMonitor)
{
var listPeptides = new List<PeptideDocNode>();
int minPrecursors = MinPrecursorsPerPeptide ?? 0;
foreach (PeptideDocNode nodePep in nodePepGroup.Children)
{
if (progressMonitor != null)
progressMonitor.ProcessMolecule(nodePep);
if (outlierIds.Contains(nodePep.Id.GlobalIndex))
continue;
// If there is a set of accepted peptides, and this is not one of them
// then skip it.
List<int> acceptedCharges = null;
if (acceptedPeptides != null &&
!acceptedPeptides.TryGetValue(AcceptModified ? new RefinementIdentity(nodePep.RawTextId) : new RefinementIdentity(nodePep.RawUnmodifiedTextId), out acceptedCharges))
{
continue;
}
int bestResultIndex = (UseBestResult ? nodePep.BestResult : -1);
float? peakFoundRatio = nodePep.GetPeakCountRatio(bestResultIndex);
if (!peakFoundRatio.HasValue)
{
if (RemoveMissingResults)
continue;
}
else
{
if (MinPeakFoundRatio.HasValue)
{
if (peakFoundRatio < MinPeakFoundRatio.Value)
continue;
}
if (MaxPeakFoundRatio.HasValue)
{
if (peakFoundRatio > MaxPeakFoundRatio.Value)
continue;
}
}
PeptideDocNode nodePepRefined = nodePep;
if (AutoPickPrecursorsAll && nodePep.AutoManageChildren == AutoPickChildrenOff)
{
nodePepRefined = (PeptideDocNode) nodePepRefined.ChangeAutoManageChildren(!AutoPickChildrenOff);
var settings = document.Settings;
if (!settings.TransitionSettings.Filter.AutoSelect && !AutoPickChildrenOff)
settings = settings.ChangeTransitionFilter(filter => filter.ChangeAutoSelect(!AutoPickChildrenOff));
nodePepRefined = nodePepRefined.ChangeSettings(settings,
new SrmSettingsDiff(false, false, true, false, AutoPickTransitionsAll, false));
}
nodePepRefined = Refine(nodePepRefined, document, bestResultIndex, acceptedCharges);
// Always remove peptides if all precursors have been removed by refinement
if (!ReferenceEquals(nodePep, nodePepRefined) && nodePepRefined.Children.Count == 0)
continue;
if (nodePepRefined.Children.Count < minPrecursors)
continue;
if (includedPeptides != null)
{
var identity = nodePepRefined.Peptide.IsCustomIon
? new RefinementIdentity(nodePep.Peptide.CustomIon)
: new RefinementIdentity(document.Settings.GetModifiedSequence(nodePepRefined));
// Skip peptides already added
if (includedPeptides.Contains(identity))
{
// Record repeated peptides for removing duplicate peptides later
if (repeatedPeptides != null)
repeatedPeptides.Add(identity);
continue;
}
// Record all peptides seen
includedPeptides.Add(identity);
}
listPeptides.Add(nodePepRefined);
}
if (MaxPepPeakRank.HasValue)
{
// Calculate the average peak area for each peptide
int countPeps = listPeptides.Count;
var listAreaIndexes = new List<PepAreaSortInfo>();
var internalStandardTypes = document.Settings.PeptideSettings.Modifications.InternalStandardTypes;
for (int i = 0; i < countPeps; i++)
{
var nodePep = listPeptides[i];
// Only peptides with children can possible be ranked by area
// Those without should be removed by this operation
if (nodePep.Children.Count == 0)
continue;
int bestResultIndex = (UseBestResult ? nodePep.BestResult : -1);
var sortInfo = new PepAreaSortInfo(nodePep, internalStandardTypes, bestResultIndex, listAreaIndexes.Count);
listAreaIndexes.Add(sortInfo);
}
listAreaIndexes.Sort((p1, p2) => Comparer.Default.Compare(p2.Area, p1.Area));
// Store area ranks
var arrayAreaIndexes = new PepAreaSortInfo[listAreaIndexes.Count];
int iRank = 1;
foreach (var areaIndex in listAreaIndexes)
{
areaIndex.Rank = iRank++;
arrayAreaIndexes[areaIndex.Index] = areaIndex;
}
// Add back all peptides with low enough rank.
listPeptides.Clear();
foreach (var areaIndex in arrayAreaIndexes)
{
if (areaIndex.Area == 0 || areaIndex.Rank > MaxPepPeakRank.Value)
continue;
listPeptides.Add(areaIndex.Peptide);
}
}
// Change the children, but only change auto-management, if the child
// identities have changed, not if their contents changed.
var childrenNew = listPeptides.ToArray();
bool updateAutoManage = !PeptideGroupDocNode.AreEquivalentChildren(nodePepGroup.Children, childrenNew);
return (PeptideGroupDocNode)nodePepGroup.ChangeChildrenChecked(childrenNew, updateAutoManage);
}
private void FinishLoad(string documentPath, MeasuredResults resultsLoad, bool changeSets)
{
if (resultsLoad == null)
{
// Loading was cancelled
_manager.EndProcessing(_document);
return;
}
SrmDocument docNew, docCurrent;
do
{
docCurrent = _container.Document;
var results = docCurrent.Settings.MeasuredResults;
// If current document has no results, then cancel
if (results == null)
{
CancelLoad(resultsLoad);
return;
}
try
{
using (var settingsChangeMonitor = new SrmSettingsChangeMonitor(new LoadMonitor(_manager, _container, null),
Resources.Loader_FinishLoad_Updating_peak_statistics,
_container, docCurrent))
{
if (changeSets)
{
// Result is empty cache, due to cancelation
results = resultsLoad.Chromatograms.Count != 0 ? resultsLoad : null;
docNew = docCurrent.ChangeMeasuredResults(results, settingsChangeMonitor);
}
else
{
// Otherwise, switch to new cache
results = results.UpdateCaches(documentPath, resultsLoad);
docNew = docCurrent.ChangeMeasuredResults(results, settingsChangeMonitor);
}
}
}
catch (OperationCanceledException)
{
// Restart the processing form the top
docNew = null;
}
}
while (docNew == null || !_manager.CompleteProcessing(_container, docNew, docCurrent));
// Force a document changed event to keep progressive load going
// until it is complete
_manager.OnDocumentChanged(_container, new DocumentChangedEventArgs(docCurrent));
}
public SrmDocument Refine(SrmDocument document, SrmSettingsChangeMonitor progressMonitor)
{
HashSet<int> outlierIds = new HashSet<int>();
if (RTRegressionThreshold.HasValue)
{
// TODO: Move necessary code into Model.
var outliers = RTLinearRegressionGraphPane.CalcOutliers(document,
RTRegressionThreshold.Value, RTRegressionPrecision, UseBestResult);
foreach (var nodePep in outliers)
outlierIds.Add(nodePep.Id.GlobalIndex);
}
HashSet<RefinementIdentity> includedPeptides = (RemoveRepeatedPeptides ? new HashSet<RefinementIdentity>() : null);
HashSet<RefinementIdentity> repeatedPeptides = (RemoveDuplicatePeptides ? new HashSet<RefinementIdentity>() : null);
Dictionary<RefinementIdentity, List<int>> acceptedPeptides = null;
if (AcceptedPeptides != null)
{
acceptedPeptides = new Dictionary<RefinementIdentity, List<int>>();
foreach (var peptideCharge in AcceptedPeptides)
{
List<int> charges;
if (!acceptedPeptides.TryGetValue(new RefinementIdentity(peptideCharge.Sequence), out charges))
{
charges = (peptideCharge.Charge.HasValue ? new List<int> {peptideCharge.Charge.Value} : null);
acceptedPeptides.Add(new RefinementIdentity(peptideCharge.Sequence), charges);
}
else if (charges != null)
{
if (peptideCharge.Charge.HasValue)
charges.Add(peptideCharge.Charge.Value);
else
acceptedPeptides[new RefinementIdentity(peptideCharge.Sequence)] = null;
}
}
}
HashSet<string> acceptedProteins = (AcceptedProteins != null ? new HashSet<string>(AcceptedProteins) : null);
var listPepGroups = new List<PeptideGroupDocNode>();
// Excluding proteins with too few peptides, since they can impact results
// of the duplicate peptide check.
int minPeptides = MinPeptidesPerProtein ?? 0;
foreach (PeptideGroupDocNode nodePepGroup in document.Children)
{
if (progressMonitor != null)
progressMonitor.ProcessGroup(nodePepGroup);
if (acceptedProteins != null && !acceptedProteins.Contains(GetAcceptProteinKey(nodePepGroup)))
continue;
PeptideGroupDocNode nodePepGroupRefined = nodePepGroup;
// If auto-managing all peptides, make sure this flag is set correctly,
// and update the peptides list, if necessary.
if (AutoPickPeptidesAll && nodePepGroup.AutoManageChildren == AutoPickChildrenOff)
{
nodePepGroupRefined =
(PeptideGroupDocNode) nodePepGroupRefined.ChangeAutoManageChildren(!AutoPickChildrenOff);
var settings = document.Settings;
if (!AutoPickChildrenOff && !settings.PeptideSettings.Filter.AutoSelect)
settings = settings.ChangePeptideFilter(filter => filter.ChangeAutoSelect(true));
nodePepGroupRefined = nodePepGroupRefined.ChangeSettings(settings,
new SrmSettingsDiff(true, false, false, false, false, false));
}
nodePepGroupRefined = Refine(nodePepGroupRefined, document, outlierIds,
includedPeptides, repeatedPeptides, acceptedPeptides, progressMonitor);
if (nodePepGroupRefined.Children.Count < minPeptides)
continue;
listPepGroups.Add(nodePepGroupRefined);
}
// Need a second pass, if all duplicate peptides should be removed,
// and duplicates were found.
if (repeatedPeptides != null && repeatedPeptides.Count > 0)
{
var listPepGroupsFiltered = new List<PeptideGroupDocNode>();
foreach (PeptideGroupDocNode nodePepGroup in listPepGroups)
{
var listPeptides = new List<PeptideDocNode>();
foreach (PeptideDocNode nodePep in nodePepGroup.Children)
{
var identity = nodePep.Peptide.IsCustomIon
? new RefinementIdentity(nodePep.Peptide.CustomIon)
: new RefinementIdentity(document.Settings.GetModifiedSequence(nodePep));
if (!repeatedPeptides.Contains(identity))
listPeptides.Add(nodePep);
}
PeptideGroupDocNode nodePepGroupRefined = (PeptideGroupDocNode)
nodePepGroup.ChangeChildrenChecked(listPeptides.ToArray(), true);
if (nodePepGroupRefined.Children.Count < minPeptides)
continue;
listPepGroupsFiltered.Add(nodePepGroupRefined);
}
listPepGroups = listPepGroupsFiltered;
}
return (SrmDocument) document.ChangeChildrenChecked(listPepGroups.ToArray(), true);
}
