public void TestChargeStateSplit(string path)
{
var data = File.ReadAllLines(path);
var map = new Dictionary <int, List <MSFeatureLight> >();
for (var i = 1; i < data.Length; i++)
{
var feature = new MSFeatureLight();
var msFeatureData = data[i].Split(',');
feature.ChargeState = Convert.ToInt32(msFeatureData[0]);
feature.MassMonoisotopic = Convert.ToDouble(msFeatureData[1]);
feature.Scan = Convert.ToInt32(msFeatureData[2]);
feature.Abundance = Convert.ToInt64(msFeatureData[3]);
if (!map.ContainsKey(feature.ChargeState))
{
map.Add(feature.ChargeState, new List <MSFeatureLight>());
}
map[feature.ChargeState].Add(feature);
}
var features = new List <UMCLight>();
foreach (var charge in map.Keys)
{
var feature = new UMCLight();
foreach (var msFeature in map[charge])
{
feature.AddChildFeature(msFeature);
}
feature.CalculateStatistics(ClusterCentroidRepresentation.Median);
features.Add(feature);
}
var finder = new MsFeatureTreeClusterer <MSFeatureLight, UMCLight>();
var comparison = finder.CompareMonoisotopic(features[0], features[1]);
Assert.AreNotEqual(comparison, 0);
}
public void TestChargeStateSplit(string path)
{
var data = File.ReadAllLines(path);
var map = new Dictionary<int, List<MSFeatureLight>>();
for (var i = 1; i < data.Length; i++)
{
var feature = new MSFeatureLight();
var msFeatureData = data[i].Split(',');
feature.ChargeState = Convert.ToInt32(msFeatureData[0]);
feature.MassMonoisotopic = Convert.ToDouble(msFeatureData[1]);
feature.Scan = Convert.ToInt32(msFeatureData[2]);
feature.Abundance = Convert.ToInt64(msFeatureData[3]);
if (!map.ContainsKey(feature.ChargeState))
{
map.Add(feature.ChargeState, new List<MSFeatureLight>());
}
map[feature.ChargeState].Add(feature);
}
var features = new List<UMCLight>();
foreach (var charge in map.Keys)
{
var feature = new UMCLight();
foreach (var msFeature in map[charge])
{
feature.AddChildFeature(msFeature);
}
feature.CalculateStatistics(ClusterCentroidRepresentation.Median);
features.Add(feature);
}
var finder = new MsFeatureTreeClusterer<MSFeatureLight, UMCLight>();
var comparison = finder.CompareMonoisotopic(features[0], features[1]);
Assert.AreNotEqual(comparison, 0);
}
public void CreateFeaturesTest(string relativePath, string outputPath)
{
var path = GetPath(relativePath);
var tolerances = new FeatureTolerances
{
Mass = 13,
Net = .01,
DriftTime = 30,
FragmentationWindowSize = .5
};
var reader = new MsFeatureLightFileReader();
var rawFeatures = reader.ReadFile(path);
var msFilterOptions = new MsFeatureFilteringOptions
{
ChargeRange = new FilterRange(1,6),
MinimumIntensity = 200000,
ShouldUseDeisotopingFilter = true,
ShouldUseIntensityFilter = true
};
rawFeatures = LcmsFeatureFilters.FilterMsFeatures(rawFeatures, msFilterOptions);
var finder = new MsFeatureTreeClusterer<MSFeatureLight, UMCLight> {Tolerances = tolerances};
finder.Progress += (sender, args) => Console.WriteLine(args.Message);
var features = finder.Cluster(rawFeatures.ToList());
var filterOptions = new LcmsFeatureFilteringOptions
{
FeatureLengthRange = new FilterRange
{
Maximum = 30,
Minimum = 10
}
};
features = LcmsFeatureFilters.FilterFeatures(features, filterOptions);
Console.WriteLine(@"Found - {0} features", features.Count);
using (var writer = File.CreateText(GetPath(outputPath)))
{
var index = 0;
foreach (var feature in features)
{
feature.Id = index++;
feature.CalculateStatistics(ClusterCentroidRepresentation.Mean);
writer.WriteLine("{1}{0}{2}{0}{3}{0}{4}{0}{5}{0}{6}{0}{7}{0}{8}{0}{9}{0}{10}",
TextDelimiter,
feature.Net,
feature.ChargeState,
feature.Mz,
feature.Scan,
feature.MassMonoisotopic,
feature.MassMonoisotopicAligned,
feature.Id,
feature.ScanStart,
feature.ScanEnd,
feature.ScanAligned
);
}
}
}
/// <summary>
/// Finds features
/// </summary>
/// <returns></returns>
public List<UMCLight> FindFeatures(List<MSFeatureLight> msFeatures,
LcmsFeatureFindingOptions options,
ISpectraProvider provider)
{
var clusterer = new MsFeatureTreeClusterer<MSFeatureLight, UMCLight>
{
Tolerances =
new FeatureTolerances
{
Mass = options.InstrumentTolerances.Mass,
Net = options.MaximumNetRange
},
ScanTolerance = options.MaximumScanRange,
SpectraProvider = provider
//TODO: Make sure we have a mass range for XIC's too....
};
clusterer.SpectraProvider = provider;
OnStatus("Starting cluster definition");
clusterer.Progress += (sender, args) => OnStatus(args.Message);
var features = clusterer.Cluster(msFeatures);
var minScan = int.MaxValue;
var maxScan = int.MinValue;
foreach (var feature in msFeatures)
{
minScan = Math.Min(feature.Scan, minScan);
maxScan = Math.Max(feature.Scan, maxScan);
}
var id = 0;
var newFeatures = new List<UMCLight>();
foreach (var feature in features)
{
if (feature.MsFeatures.Count < 1)
continue;
feature.Net = Convert.ToDouble(feature.Scan - minScan)/Convert.ToDouble(maxScan - minScan);
feature.CalculateStatistics(ClusterCentroidRepresentation.Median);
feature.Net = feature.Net;
feature.Id = id++;
newFeatures.Add(feature);
//Sets the width of the feature to be the width of the peak, not the width of the tails
var maxAbundance = double.MinValue;
var maxAbundanceIndex = 0;
for (var msFeatureIndex = 0; msFeatureIndex < feature.MsFeatures.Count - 1; msFeatureIndex++)
{
var msFeature = feature.MsFeatures[msFeatureIndex];
if (msFeature.Abundance > maxAbundance)
{
maxAbundance = msFeature.Abundance;
maxAbundanceIndex = msFeatureIndex;
}
}
for (var msFeatureIndex = maxAbundanceIndex; msFeatureIndex > 0; msFeatureIndex--)
{
if (feature.MsFeatures[msFeatureIndex].Abundance / maxAbundance <= 0.05)
{
feature.ScanStart = feature.MsFeatures[msFeatureIndex].Scan;
break;
}
}
for (var msFeatureIndex = maxAbundanceIndex; msFeatureIndex < feature.MsFeatures.Count - 1; msFeatureIndex++)
{
if (feature.MsFeatures[msFeatureIndex].Abundance / maxAbundance <= 0.05)
{
feature.ScanEnd = feature.MsFeatures[msFeatureIndex].Scan;
break;
}
}
}
return features;
}
public void CreateFeaturesTest(string relativePath, string outputPath)
{
var path = GetPath(relativePath);
var tolerances = new FeatureTolerances
{
Mass = 13,
Net = .01,
DriftTime = 30,
FragmentationWindowSize = .5
};
var reader = new MsFeatureLightFileReader();
var rawFeatures = reader.ReadFile(path);
var msFilterOptions = new MsFeatureFilteringOptions
{
ChargeRange = new FilterRange(1, 6),
MinimumIntensity = 200000,
ShouldUseDeisotopingFilter = true,
ShouldUseIntensityFilter = true
};
rawFeatures = LcmsFeatureFilters.FilterMsFeatures(rawFeatures, msFilterOptions);
var finder = new MsFeatureTreeClusterer <MSFeatureLight, UMCLight> {
Tolerances = tolerances
};
finder.Progress += (sender, args) => Console.WriteLine(args.Message);
var features = finder.Cluster(rawFeatures.ToList());
var filterOptions = new LcmsFeatureFilteringOptions
{
FeatureLengthRangeScans = new FilterRange
{
Maximum = 30,
Minimum = 10
}
};
features = LcmsFeatureFilters.FilterFeatures(features, filterOptions);
Console.WriteLine(@"Found - {0} features", features.Count);
using (var writer = File.CreateText(GetPath(outputPath)))
{
var index = 0;
foreach (var feature in features)
{
feature.Id = index++;
feature.CalculateStatistics(ClusterCentroidRepresentation.Mean);
writer.WriteLine("{1}{0}{2}{0}{3}{0}{4}{0}{5}{0}{6}{0}{7}{0}{8}{0}{9}{0}{10}",
TextDelimiter,
feature.Net,
feature.ChargeState,
feature.Mz,
feature.Scan,
feature.MassMonoisotopic,
feature.MassMonoisotopicAligned,
feature.Id,
feature.ScanStart,
feature.ScanEnd,
feature.ScanAligned
);
}
}
}