public void TestClusterWriting(string databasePath, string crossPath, int charge, int minimumClusterSize)
{
NHibernateUtil.ConnectToDatabase(databasePath, false);
IDatasetDAO datasetCache = new DatasetDAOHibernate();
IUmcClusterDAO clusterCache = new UmcClusterDAOHibernate();
IUmcDAO featureCache = new UmcDAOHibernate();
Console.WriteLine("Find all datasets");
var datasets = datasetCache.FindAll();
Console.WriteLine("Find all clusters");
var clusters = clusterCache.FindByCharge(charge);
WriteClusters(datasets, clusters, minimumClusterSize, charge, crossPath, databasePath, 300000);
}
public void CreateCrossTab(string databasePath, string crossPath, int charge, int minimumClusterSize)
{
NHibernateUtil.ConnectToDatabase(databasePath, false);
IDatasetDAO datasetCache = new DatasetDAOHibernate();
IUmcClusterDAO clusterCache = new UmcClusterDAOHibernate();
IUmcDAO featureCache = new UmcDAOHibernate();
Console.WriteLine("Find all datasets");
var datasets = datasetCache.FindAll();
Console.WriteLine("Find all clusters");
var clusters = clusterCache.FindByCharge(charge);
WriteClusters(datasets, clusters, minimumClusterSize, charge, crossPath, databasePath, 50000);
}
public void CreateDatasetMap(string databasePath, string crossPath, int charge, int minimumClusterSize)
{
NHibernateUtil.ConnectToDatabase(databasePath, false);
IDatasetDAO datasetCache = new DatasetDAOHibernate();
IUmcClusterDAO clusterCache = new UmcClusterDAOHibernate();
IUmcDAO featureCache = new UmcDAOHibernate();
Console.WriteLine("Find all datasets");
var datasets = datasetCache.FindAll();
using (TextWriter writer = File.CreateText(crossPath + ".csv"))
{
writer.WriteLine("Dataset, Dataset Id");
foreach (var info in datasets)
{
writer.WriteLine("{0},{1}", info.DatasetName, info.DatasetId);
}
}
}
public void CreateDatasetMap(string databasePath, string crossPath, int charge, int minimumClusterSize)
{
NHibernateUtil.ConnectToDatabase(databasePath, false);
IDatasetDAO datasetCache = new DatasetDAOHibernate();
IUmcClusterDAO clusterCache = new UmcClusterDAOHibernate();
IUmcDAO featureCache = new UmcDAOHibernate();
Console.WriteLine("Find all datasets");
var datasets = datasetCache.FindAll();
using (TextWriter writer = File.CreateText(crossPath + ".csv"))
{
writer.WriteLine("Dataset, Dataset Id");
foreach (var info in datasets)
{
writer.WriteLine("{0},{1}", info.DatasetName, info.DatasetId);
}
}
}
/// <summary>
/// The main entry point for the application.
/// </summary>
static int Main(string [] args)
{
var handle = System.Diagnostics.Process.GetCurrentProcess().MainWindowHandle;
SetConsoleMode(handle, ENABLE_EXTENDED_FLAGS);
try
{
if (args.Length < 2)
{
Console.WriteLine(@"MultiAlignChargeStateProcessor databasePath chargeState crossTabPath [dataset List]");
Console.WriteLine(@"\tThe cross-tab file will be placed in the same directory as the database path");
return(1);
}
// Setup the analysis processing
var databasePath = args[0];
var databaseName = Path.GetFileNameWithoutExtension(databasePath);
var path = Path.GetDirectoryName(databasePath);
var crossPath = args[2];
var chargeState = Convert.ToInt32(args[1]);
List <string> datasetList = null;
if (args.Length == 4)
{
datasetList = File.ReadAllLines(args[3]).ToList();
}
if (path == null)
{
Console.WriteLine(@"The directory path is invalid");
return(1);
}
NHibernateUtil.ConnectToDatabase(databasePath, false);
IDatasetDAO datasetCache = new DatasetDAOHibernate();
var dateSuffix = AnalysisPathUtils.BuildDateSuffix();
Logger.LogPath = Path.Combine(path, string.Format("{0}_charge_{2}_{1}.txt", databaseName, dateSuffix, chargeState));
Logger.PrintMessage("Find all datasets", true);
var datasets = datasetCache.FindAll();
Logger.PrintMessage(string.Format("Found {0} datasets", datasets.Count), true);
// Create the clustering algorithm - average linkage
IClusterer <UMCLight, UMCClusterLight> clusterer = new UMCAverageLinkageClusterer <UMCLight, UMCClusterLight>();
// Create the DAO object to extract the features
var database = new UmcAdoDAO {
DatabasePath = databasePath
};
IUmcDAO featureDao = database;
Logger.PrintMessage(string.Format("Extracting Features"), true);
var tempFeatures = featureDao.FindByCharge(chargeState);
Logger.PrintMessage(string.Format("Found {0} features", tempFeatures.Count), true);
var features = new List <UMCLight>();
if (datasetList != null)
{
var featuremap = datasets.ToDictionary(info => info.DatasetName.ToLower());
var focusedDatasetList = new Dictionary <int, DatasetInformation>();
foreach (var name in datasetList)
{
var key = name.ToLower();
if (featuremap.ContainsKey(key))
{
Logger.PrintMessage("Using dataset: " + name);
focusedDatasetList.Add(featuremap[key].DatasetId, featuremap[key]);
}
else
{
throw new Exception("Didn't find the dataset required..." + name);
}
}
features.AddRange(from feature in tempFeatures let use = focusedDatasetList.ContainsKey(feature.GroupId) where use select feature);
Logger.PrintMessage(string.Format("Found {0} filtered features for dataset list", features.Count), true);
}
else
{
features = tempFeatures;
}
// Handle logging progress.
clusterer.Progress += clusterer_Progress;
clusterer.Parameters.Tolerances.DriftTime = .3;
clusterer.Parameters.Tolerances.Mass = 16;
clusterer.Parameters.Tolerances.Net = .014;
clusterer.Parameters.OnlyClusterSameChargeStates = true;
clusterer.Parameters.CentroidRepresentation = ClusterCentroidRepresentation.Mean;
clusterer.Parameters.DistanceFunction = DistanceFactory <UMCLight> .CreateDistanceFunction(DistanceMetric.WeightedEuclidean);
// Then cluster
var clusterWriter = new UmcClusterWriter();
IClusterWriter <UMCClusterLight> writer = clusterWriter; //new UMCClusterDummyWriter();
try
{
clusterWriter.Open(crossPath);
clusterWriter.WriteHeader(datasets);
clusterer.ClusterAndProcess(features, writer);
Logger.PrintMessage("", true);
Logger.PrintMessage("ANALYSIS SUCCESS", true);
return(0);
}
catch (Exception ex)
{
Logger.PrintMessage("Unhandled Error: " + ex.Message);
var innerEx = ex.InnerException;
while (innerEx != null)
{
Logger.PrintMessage("Inner Exception: " + innerEx.Message);
innerEx = innerEx.InnerException;
}
Logger.PrintMessage("Stack: " + ex.StackTrace);
Logger.PrintMessage("");
Logger.PrintMessage("ANALYSIS FAILED");
return(1);
}
finally
{
clusterWriter.Close();
}
}
catch (Exception ex)
{
Logger.PrintMessage("Unhandled Error: " + ex.Message, true);
var innerEx = ex.InnerException;
while (innerEx != null)
{
Logger.PrintMessage("Inner Exception: " + innerEx.Message);
innerEx = innerEx.InnerException;
}
Logger.PrintMessage("Stack: " + ex.StackTrace, true);
Logger.PrintMessage("");
Logger.PrintMessage("ANALYSIS FAILED");
return(1);
}
}
public void TestClusterGeneration(string databasePath,
string crossPath,
int charge,
int minimumClusterSize)
{
File.Delete(databasePath);
NHibernateUtil.ConnectToDatabase(databasePath, true);
IDatasetDAO datasetCache = new DatasetDAOHibernate();
IUmcClusterDAO clusterCache = new UmcClusterDAOHibernate();
IUmcDAO featureCache = new UmcDAOHibernate();
// Creating a dataset
Console.WriteLine("Creating dummy datasets");
var datasets = new List<DatasetInformation>();
var total = 10;
for (var i = 0; i < total; i++)
{
var dataset = new DatasetInformation();
dataset.DatasetId = i;
dataset.DatasetName = "test" + i;
datasets.Add(dataset);
}
datasetCache.AddAll(datasets);
datasets.Clear();
datasets = datasetCache.FindAll();
// Create features
Console.WriteLine("Creating features");
var features = new List<UMCLight>();
var clusters = new List<UMCClusterLight>();
var x = new Random();
var featureId = 0;
for (var i = 0; i < 100; i++)
{
var cluster = new UMCClusterLight();
cluster.Id = i;
cluster.AmbiguityScore = i;
cluster.Tightness = i;
var N = x.Next(1, total);
cluster.Id = i;
cluster.ChargeState = charge;
var hash = new HashSet<int>();
for (var j = 0; j < N; j++)
{
var did = -1;
do
{
did = x.Next(0, total);
if (!hash.Contains(did))
{
hash.Add(did);
break;
}
} while (true);
var feature = new UMCLight();
feature.GroupId = did;
feature.Id = featureId++;
feature.ChargeState = charge;
feature.MassMonoisotopic = x.NextDouble();
feature.Net = x.NextDouble();
feature.AbundanceSum = x.Next(100, 200);
feature.Abundance = feature.Abundance;
feature.ClusterId = cluster.Id;
cluster.AddChildFeature(feature);
features.Add(feature);
}
cluster.CalculateStatistics(ClusterCentroidRepresentation.Mean);
clusters.Add(cluster);
}
featureCache.AddAll(features);
clusterCache.AddAll(clusters);
clusters = clusterCache.FindAll();
Console.WriteLine("Find all clusters");
clusters = clusterCache.FindByCharge(charge);
WriteClusters(datasets,
clusters,
minimumClusterSize,
charge,
crossPath,
databasePath,
300000);
}
public void TestCreateDummyDatabase(string databasePath, int totalDatasets, int totalClusters)
{
File.Delete(databasePath);
NHibernateUtil.ConnectToDatabase(databasePath, true);
IDatasetDAO datasetCache = new DatasetDAOHibernate();
IUmcClusterDAO clusterCache = new UmcClusterDAOHibernate();
IUmcDAO featureCache = new UmcDAOHibernate();
// Creating a dataset
Console.WriteLine("Creating dummy datasets");
var datasets = new List<DatasetInformation>();
var total = totalDatasets;
for (var i = 0; i < total; i++)
{
var dataset = new DatasetInformation();
dataset.DatasetId = i;
dataset.DatasetName = "test" + i;
datasets.Add(dataset);
}
datasetCache.AddAll(datasets);
datasets.Clear();
datasets = datasetCache.FindAll();
// Create features
Console.WriteLine("Creating features");
var features = new List<UMCLight>();
var clusters = new List<UMCClusterLight>();
var x = new Random();
var featureId = 0;
for (var i = 0; i < totalClusters; i++)
{
var N = x.Next(1, total);
var charge = x.Next(1, 10);
var hash = new HashSet<int>();
var net = x.NextDouble();
var mass = 400 + (1600*x.NextDouble());
var dt = 60*x.NextDouble();
for (var j = 0; j < N; j++)
{
var did = -1;
do
{
did = x.Next(0, total);
if (!hash.Contains(did))
{
hash.Add(did);
break;
}
} while (true);
var feature = new UMCLight
{
GroupId = did,
Id = featureId++,
ChargeState = charge,
MassMonoisotopic = FeatureLight.ComputeDaDifferenceFromPPM(mass, 3)
};
feature.MassMonoisotopicAligned = feature.MassMonoisotopic;
feature.Net = net + 0.03 * x.NextDouble();
feature.NetAligned = feature.Net;
feature.Net = feature.Net;
feature.DriftTime = dt;
feature.AbundanceSum = x.Next(100, 200);
feature.Abundance = feature.Abundance;
feature.ClusterId = -1;
features.Add(feature);
}
}
featureCache.AddAll(features);
}
public void TestCreateDummyDatabase(string databasePath, int totalDatasets, int totalClusters)
{
File.Delete(databasePath);
NHibernateUtil.ConnectToDatabase(databasePath, true);
IDatasetDAO datasetCache = new DatasetDAOHibernate();
IUmcClusterDAO clusterCache = new UmcClusterDAOHibernate();
IUmcDAO featureCache = new UmcDAOHibernate();
// Creating a dataset
Console.WriteLine("Creating dummy datasets");
var datasets = new List <DatasetInformation>();
var total = totalDatasets;
for (var i = 0; i < total; i++)
{
var dataset = new DatasetInformation();
dataset.DatasetId = i;
dataset.DatasetName = "test" + i;
datasets.Add(dataset);
}
datasetCache.AddAll(datasets);
datasets.Clear();
datasets = datasetCache.FindAll();
// Create features
Console.WriteLine("Creating features");
var features = new List <UMCLight>();
var clusters = new List <UMCClusterLight>();
var x = new Random();
var featureId = 0;
for (var i = 0; i < totalClusters; i++)
{
var N = x.Next(1, total);
var charge = x.Next(1, 10);
var hash = new HashSet <int>();
var net = x.NextDouble();
var mass = 400 + (1600 * x.NextDouble());
var dt = 60 * x.NextDouble();
for (var j = 0; j < N; j++)
{
var did = -1;
do
{
did = x.Next(0, total);
if (!hash.Contains(did))
{
hash.Add(did);
break;
}
} while (true);
var feature = new UMCLight();
feature.GroupId = did;
feature.Id = featureId++;
feature.ChargeState = charge;
feature.MassMonoisotopic = FeatureLight.ComputeDaDifferenceFromPPM(mass, 3);
feature.MassMonoisotopicAligned = feature.MassMonoisotopic;
feature.Net = net + .03 * x.NextDouble();
feature.NetAligned = feature.Net;
feature.Net = feature.Net;
feature.DriftTime = dt;
feature.AbundanceSum = x.Next(100, 200);
feature.Abundance = feature.Abundance;
feature.ClusterId = -1;
features.Add(feature);
}
}
featureCache.AddAll(features);
}
public void TestClusterGeneration(string databasePath,
string crossPath,
int charge,
int minimumClusterSize)
{
File.Delete(databasePath);
NHibernateUtil.ConnectToDatabase(databasePath, true);
IDatasetDAO datasetCache = new DatasetDAOHibernate();
IUmcClusterDAO clusterCache = new UmcClusterDAOHibernate();
IUmcDAO featureCache = new UmcDAOHibernate();
// Creating a dataset
Console.WriteLine("Creating dummy datasets");
var datasets = new List <DatasetInformation>();
var total = 10;
for (var i = 0; i < total; i++)
{
var dataset = new DatasetInformation();
dataset.DatasetId = i;
dataset.DatasetName = "test" + i;
datasets.Add(dataset);
}
datasetCache.AddAll(datasets);
datasets.Clear();
datasets = datasetCache.FindAll();
// Create features
Console.WriteLine("Creating features");
var features = new List <UMCLight>();
var clusters = new List <UMCClusterLight>();
var x = new Random();
var featureId = 0;
for (var i = 0; i < 100; i++)
{
var cluster = new UMCClusterLight();
cluster.Id = i;
cluster.AmbiguityScore = i;
cluster.Tightness = i;
var N = x.Next(1, total);
cluster.Id = i;
cluster.ChargeState = charge;
var hash = new HashSet <int>();
for (var j = 0; j < N; j++)
{
var did = -1;
do
{
did = x.Next(0, total);
if (!hash.Contains(did))
{
hash.Add(did);
break;
}
} while (true);
var feature = new UMCLight();
feature.GroupId = did;
feature.Id = featureId++;
feature.ChargeState = charge;
feature.MassMonoisotopic = x.NextDouble();
feature.Net = x.NextDouble();
feature.AbundanceSum = x.Next(100, 200);
feature.Abundance = feature.Abundance;
feature.ClusterId = cluster.Id;
cluster.AddChildFeature(feature);
features.Add(feature);
}
cluster.CalculateStatistics(ClusterCentroidRepresentation.Mean);
clusters.Add(cluster);
}
featureCache.AddAll(features);
clusterCache.AddAll(clusters);
clusters = clusterCache.FindAll();
Console.WriteLine("Find all clusters");
clusters = clusterCache.FindByCharge(charge);
WriteClusters(datasets,
clusters,
minimumClusterSize,
charge,
crossPath,
databasePath,
300000);
}
/// <summary>
/// The main entry point for the application.
/// </summary>
static int Main(string [] args)
{
var handle = System.Diagnostics.Process.GetCurrentProcess().MainWindowHandle;
SetConsoleMode(handle, ENABLE_EXTENDED_FLAGS);
try
{
if (args.Length < 2)
{
Console.WriteLine(@"MultiAlignChargeStateProcessor databasePath chargeState crossTabPath [dataset List]");
Console.WriteLine(@"\tThe cross-tab file will be placed in the same directory as the database path");
return 1;
}
// Setup the analysis processing
var databasePath = args[0];
var databaseName = Path.GetFileNameWithoutExtension(databasePath);
var path = Path.GetDirectoryName(databasePath);
var crossPath = args[2];
var chargeState = Convert.ToInt32(args[1]);
List<string> datasetList = null;
if (args.Length == 4)
{
datasetList = File.ReadAllLines(args[3]).ToList();
}
if (path == null)
{
Console.WriteLine(@"The directory path is invalid");
return 1;
}
NHibernateUtil.ConnectToDatabase(databasePath, false);
IDatasetDAO datasetCache = new DatasetDAOHibernate();
var dateSuffix = AnalysisPathUtils.BuildDateSuffix();
Logger.LogPath = Path.Combine(path, string.Format("{0}_charge_{2}_{1}.txt", databaseName, dateSuffix, chargeState));
Logger.PrintMessage("Find all datasets", true);
var datasets = datasetCache.FindAll();
Logger.PrintMessage(string.Format("Found {0} datasets", datasets.Count), true);
// Create the clustering algorithm - average linkage
IClusterer<UMCLight, UMCClusterLight> clusterer = new UMCAverageLinkageClusterer<UMCLight, UMCClusterLight>();
// Create the DAO object to extract the features
var database = new UmcAdoDAO {DatabasePath = databasePath};
IUmcDAO featureDao = database;
Logger.PrintMessage(string.Format("Extracting Features"), true);
var tempFeatures = featureDao.FindByCharge(chargeState);
Logger.PrintMessage(string.Format("Found {0} features", tempFeatures.Count), true);
var features = new List<UMCLight>();
if (datasetList != null)
{
var featuremap = datasets.ToDictionary(info => info.DatasetName.ToLower());
var focusedDatasetList = new Dictionary<int, DatasetInformation>();
foreach (var name in datasetList)
{
var key = name.ToLower();
if (featuremap.ContainsKey(key))
{
Logger.PrintMessage("Using dataset: " + name);
focusedDatasetList.Add(featuremap[key].DatasetId, featuremap[key]);
}
else
throw new Exception("Didn't find the dataset required..." + name);
}
features.AddRange(from feature in tempFeatures let use = focusedDatasetList.ContainsKey(feature.GroupId) where use select feature);
Logger.PrintMessage(string.Format("Found {0} filtered features for dataset list", features.Count), true);
}
else
{
features = tempFeatures;
}
// Handle logging progress.
clusterer.Progress += clusterer_Progress;
clusterer.Parameters.Tolerances.DriftTime = .3;
clusterer.Parameters.Tolerances.Mass = 16;
clusterer.Parameters.Tolerances.Net = .014;
clusterer.Parameters.OnlyClusterSameChargeStates = true;
clusterer.Parameters.CentroidRepresentation = ClusterCentroidRepresentation.Mean;
clusterer.Parameters.DistanceFunction = PNNLOmics.Algorithms.Distance.DistanceFactory<UMCLight>.CreateDistanceFunction(PNNLOmics.Algorithms.Distance.DistanceMetric.WeightedEuclidean);
// Then cluster
var clusterWriter = new UmcClusterWriter();
IClusterWriter<UMCClusterLight> writer = clusterWriter; //new UMCClusterDummyWriter();
try
{
clusterWriter.Open(crossPath);
clusterWriter.WriteHeader(datasets);
clusterer.ClusterAndProcess(features, writer);
Logger.PrintMessage("", true);
Logger.PrintMessage("ANALYSIS SUCCESS", true);
return 0;
}
catch (Exception ex)
{
Logger.PrintMessage("Unhandled Error: " + ex.Message);
var innerEx = ex.InnerException;
while (innerEx != null)
{
Logger.PrintMessage("Inner Exception: " + innerEx.Message);
innerEx = innerEx.InnerException;
}
Logger.PrintMessage("Stack: " + ex.StackTrace);
Logger.PrintMessage("");
Logger.PrintMessage("ANALYSIS FAILED");
return 1;
}
finally
{
clusterWriter.Close();
}
}
catch (Exception ex)
{
Logger.PrintMessage("Unhandled Error: " + ex.Message, true);
var innerEx = ex.InnerException;
while (innerEx != null)
{
Logger.PrintMessage("Inner Exception: " + innerEx.Message);
innerEx = innerEx.InnerException;
}
Logger.PrintMessage("Stack: " + ex.StackTrace, true);
Logger.PrintMessage("");
Logger.PrintMessage("ANALYSIS FAILED");
return 1;
}
}