public static void OGlycoTest_Localization2()
{
//There may have a bug that MM cannot identify Peptide modified with (HexNAc), This is to test and find the bug.
//Get glycanBox
var glycanBox = OGlycanBoxes[0];
//Get unmodified peptide, products, allPossible modPos and all boxes.
Protein protein = new Protein("AATVGSLAGQPLQER", "P16150");
var peptide = protein.Digest(new DigestionParams(), new List <Modification>(), new List <Modification>()).First();
List <Product> products = new List <Product>();
peptide.Fragment(DissociationType.ETD, FragmentationTerminus.Both, products);
int[] modPos = GlycoSpectralMatch.GetPossibleModSites(peptide, new string[] { "S", "T" }).OrderBy(p => p).ToArray();
var boxes = GlycanBox.BuildChildOGlycanBoxes(glycanBox.NumberOfMods, glycanBox.ModIds).ToArray();
//Load scan.
CommonParameters commonParameters = new CommonParameters(dissociationType: DissociationType.ETD, trimMsMsPeaks: false);
string spectraFile = Path.Combine(TestContext.CurrentContext.TestDirectory, @"GlycoTestData\181217_Fusion_(LC2)_NewObj_Serum_deSA_Jacalin_HRM_4h_ETD_HCD_DDA_mz(400_1200)_21707.mgf");
var file = new MyFileManager(true).LoadFile(spectraFile, commonParameters);
var scans = MetaMorpheusTask.GetMs2Scans(file, spectraFile, commonParameters).ToArray();
//Known peptideWithMod match.
var peptideWithMod = GlycoPeptides.OGlyGetTheoreticalPeptide(new int[1] {
4
}, peptide, glycanBox);
Assert.That(peptideWithMod.FullSequence == "AAT[O-Glycosylation:N1 on X]VGSLAGQPLQER");
//List<Product> knownProducts = peptideWithMod.Fragment(DissociationType.EThcD, FragmentationTerminus.Both).ToList();
List <Product> knownProducts = GlycoPeptides.OGlyGetTheoreticalFragments(DissociationType.ETD, peptide, peptideWithMod);
var matchedKnownFragmentIons = MetaMorpheusEngine.MatchFragmentIons(scans.First(), knownProducts, commonParameters);
//Get hashset int
int obsPreviousFragmentCeilingMz = 0;
List <int> binsToSearch = new List <int>();
foreach (var envelope in scans.First().ExperimentalFragments)
{
// assume charge state 1 to calculate mass tolerance
double experimentalFragmentMass = envelope.MonoisotopicMass;
// get theoretical fragment bins within mass tolerance
int obsFragmentFloorMass = (int)Math.Floor((commonParameters.ProductMassTolerance.GetMinimumValue(experimentalFragmentMass)) * 1000);
int obsFragmentCeilingMass = (int)Math.Ceiling((commonParameters.ProductMassTolerance.GetMaximumValue(experimentalFragmentMass)) * 1000);
// prevents double-counting peaks close in m/z and lower-bound out of range exceptions
if (obsFragmentFloorMass < obsPreviousFragmentCeilingMz)
{
obsFragmentFloorMass = obsPreviousFragmentCeilingMz;
}
obsPreviousFragmentCeilingMz = obsFragmentCeilingMass + 1;
// search mass bins within a tolerance
for (int fragmentBin = obsFragmentFloorMass; fragmentBin <= obsFragmentCeilingMass; fragmentBin++)
{
binsToSearch.Add(fragmentBin);
}
}
HashSet <int> allPeaks = new HashSet <int>(binsToSearch);
//Graph Localization
LocalizationGraph localizationGraph = new LocalizationGraph(modPos, glycanBox, boxes, -1);
LocalizationGraph.LocalizeOGlycan(localizationGraph, scans.First(), commonParameters.ProductMassTolerance, products);
var allPaths = LocalizationGraph.GetAllHighestScorePaths(localizationGraph.array, localizationGraph.ChildModBoxes);
var knowPath = new int[2] {
1, 1
};
Assert.That(Enumerable.SequenceEqual(knowPath, allPaths[0]));
var local = LocalizationGraph.GetLocalizedPath(localizationGraph, allPaths.First());
Assert.That(Enumerable.SequenceEqual(local.Mods.Select(p => p.Item1), new List <int> {
4
}));
Assert.That(Enumerable.SequenceEqual(local.Mods.Select(p => p.Item2), new List <int> {
0
}));
}
public static void OGlycoTest_Localization()
{
//Get glycanBox
var glycanBox = OGlycanBoxes[19];
//Get unmodified peptide, products, allPossible modPos and all boxes.
Protein protein = new Protein("TTGSLEPSSGASGPQVSSVK", "P16150");
var peptide = protein.Digest(new DigestionParams(), new List <Modification>(), new List <Modification>()).First();
List <Product> products = new List <Product>();
peptide.Fragment(DissociationType.ETD, FragmentationTerminus.Both, products);
int[] modPos = GlycoSpectralMatch.GetPossibleModSites(peptide, new string[] { "S", "T" }).OrderBy(v => v).ToArray();
var boxes = GlycanBox.BuildChildOGlycanBoxes(3, glycanBox.ModIds).ToArray();
Assert.That(boxes.Count() == 6);
//Get Unlocal Fragment
var unlocalCost = GlycoPeptides.GetUnlocalFragment(products, modPos, glycanBox);
Assert.That(unlocalCost.Count == 4); //Basicly, the unlocal are c/z ions that don't localize glycosylation.
//Get scan
CommonParameters commonParameters = new CommonParameters(dissociationType: DissociationType.EThcD, trimMsMsPeaks: false);
string spectraFile = Path.Combine(TestContext.CurrentContext.TestDirectory, @"GlycoTestData\2019_09_16_StcEmix_35trig_EThcD25_rep1_4565.mgf");
var file = new MyFileManager(true).LoadFile(spectraFile, commonParameters);
var scans = MetaMorpheusTask.GetMs2Scans(file, spectraFile, commonParameters).ToArray();
//Known peptideWithMod match.
var peptideWithMod = GlycoPeptides.OGlyGetTheoreticalPeptide(new int[3] {
10, 2, 3
}, peptide, glycanBox);
Assert.That(peptideWithMod.FullSequence == "T[O-Glycosylation:H1N1 on X]T[O-Glycosylation:H1N1 on X]GSLEPSS[O-Glycosylation:N1 on X]GASGPQVSSVK");
List <Product> knownProducts = GlycoPeptides.OGlyGetTheoreticalFragments(DissociationType.EThcD, peptide, peptideWithMod);
var matchedKnownFragmentIons = MetaMorpheusEngine.MatchFragmentIons(scans.First(), knownProducts, commonParameters);
//Graph Localization
LocalizationGraph localizationGraph = new LocalizationGraph(modPos, glycanBox, boxes, -1);
LocalizationGraph.LocalizeOGlycan(localizationGraph, scans.First(), commonParameters.ProductMassTolerance, products);
var allPaths = LocalizationGraph.GetAllHighestScorePaths(localizationGraph.array, localizationGraph.ChildModBoxes);
var knowPath = new int[8] {
2, 4, 4, 4, 5, 5, 5, 5
};
Assert.That(Enumerable.SequenceEqual(knowPath, allPaths[0]));
//Get localized Route
var local = LocalizationGraph.GetLocalizedPath(localizationGraph, allPaths.First());
Assert.That(Enumerable.SequenceEqual(local.Mods.Select(v => v.Item1), new List <int> {
2, 3, 10
}));
Assert.That(Enumerable.SequenceEqual(local.Mods.Select(v => v.Item2), new List <int> {
1, 1, 0
}));
//Get all paths, calculate PScore and calculate position probability.
var p = scans.First().TheScan.MassSpectrum.Size *commonParameters.ProductMassTolerance.GetRange(1000).Width / scans.First().TheScan.MassSpectrum.Range.Width;
var n = knownProducts.Where(v => v.ProductType == ProductType.c || v.ProductType == ProductType.zDot).Count();
var allPathWithWeights = LocalizationGraph.GetAllPaths_CalP(localizationGraph, p, n);
Assert.That(allPathWithWeights.Count == 168);
//Calculate Site Specific Localization Probability
var y = LocalizationGraph.CalSiteSpecificLocalizationProbability(allPathWithWeights, localizationGraph.ModPos);
Assert.That(y.Count == 8);
Assert.That(y.First().Value[1].Item2 > 0.99);
}