/// <summary>
/// Validate Muscle multiple sequence alignment with static properties
/// of PamsamMultipleSequenceAligner.
/// </summary>
/// <param name="nodeName">xml node name</param>
/// <param name="hierarchicalClusteringMethodName"></param>
/// <param name="distanceFunctionName">kmerdistancematrix method name.</param>
/// <param name="profileAlignerName"></param>
/// <param name="profileScoreName">Profile score function name.</param>
/// <param name="useweights">use sequence weights true\false</param>
/// <param name="fasterVersion">fasterversion true\false</param>
/// <param name="useStageB">stage2 computation true\false</param>
/// <param name="expectedScoreNode"></param>
private void ValidatePamsamAlign(string nodeName,
string expectedScoreNode,
UpdateDistanceMethodsTypes hierarchicalClusteringMethodName,
DistanceFunctionTypes distanceFunctionName,
ProfileAlignerNames profileAlignerName,
ProfileScoreFunctionNames profileScoreName,
bool useweights,
bool fasterVersion,
bool useStageB)
{
Initialize(nodeName, expectedScoreNode);
// get old properties
bool prevVersion = PAMSAMMultipleSequenceAligner.FasterVersion;
bool prevUseWeights = PAMSAMMultipleSequenceAligner.UseWeights;
bool prevUseStageB = PAMSAMMultipleSequenceAligner.UseStageB;
try
{
// Set static properties
PAMSAMMultipleSequenceAligner.FasterVersion = fasterVersion;
PAMSAMMultipleSequenceAligner.UseWeights = useweights;
PAMSAMMultipleSequenceAligner.UseStageB = useStageB;
// MSA aligned sequences.
int numberOfDegrees = 2;
int numberOfPartitions = 2;
var msa =
new PAMSAMMultipleSequenceAligner(lstSequences,
kmerLength, distanceFunctionName, hierarchicalClusteringMethodName,
profileAlignerName, profileScoreName, similarityMatrix,
gapOpenPenalty,
gapExtendPenalty, numberOfDegrees, numberOfPartitions);
// Validate the aligned Sequence and score
if (fasterVersion)
{
InitializeStage1Variables(nodeName);
Assert.AreEqual(stage1ExpectedSequences.Count, msa.AlignedSequences.Count);
int index = 0;
foreach (ISequence seq in msa.AlignedSequences)
{
Assert.AreEqual(new string(seq.Select(a => (char) a).ToArray()),
new string(stage1ExpectedSequences[index].Select(a => (char) a).ToArray()));
index++;
}
Assert.IsTrue(stage1ExpectedScore.Contains(msa.AlignmentScore.ToString((IFormatProvider) null)));
}
else
{
int index = 0;
foreach (ISequence seq in msa.AlignedSequences)
{
Assert.AreEqual(new string(seq.Select(a => (char) a).ToArray()),
new string(expectedSequences[index].Select(a => (char) a).ToArray()));
index++;
}
Assert.AreEqual(expectedScore, msa.AlignmentScore.ToString((IFormatProvider) null));
}
}
finally
{
// Reset it back
PAMSAMMultipleSequenceAligner.FasterVersion = prevVersion;
PAMSAMMultipleSequenceAligner.UseWeights = prevUseWeights;
PAMSAMMultipleSequenceAligner.UseStageB = prevUseStageB;
}
ApplicationLog.WriteLine(
String.Format(null, @"Validation of pamsam alignment completed
successfully for molecule type {0} with
static property fasterversion {0}, usestageb {1} and useweights {2}",
fasterVersion, useStageB, useweights));
}
/// <summary>
/// Aligns multiple sequences using a multiple sequence aligner.
/// This sample uses PAMSAM with a set of default parameters.
/// </summary>
/// <param name="sequences">List of sequences to align.</param>
/// <returns>List of ISequenceAlignment</returns>
public static IList<ISequence> DoMultipleSequenceAlignment(List<ISequence> sequences)
{
// Initialize objects for constructor
SimilarityMatrix similarityMatrix = new SimilarityMatrix(SimilarityMatrix.StandardSimilarityMatrix.AmbiguousDna);
int gapOpenPenalty = -4;
int gapExtendPenalty = -1;
int kmerLength = 3;
DistanceFunctionTypes distanceFunctionName = DistanceFunctionTypes.EuclideanDistance;
UpdateDistanceMethodsTypes hierarchicalClusteringMethodName = UpdateDistanceMethodsTypes.Average;
ProfileAlignerNames profileAlignerName = ProfileAlignerNames.NeedlemanWunschProfileAligner;
ProfileScoreFunctionNames profileProfileFunctionName = ProfileScoreFunctionNames.WeightedInnerProduct;
// Call aligner
PAMSAMMultipleSequenceAligner msa = new PAMSAMMultipleSequenceAligner
(sequences, kmerLength, distanceFunctionName, hierarchicalClusteringMethodName,
profileAlignerName, profileProfileFunctionName, similarityMatrix, gapOpenPenalty, gapExtendPenalty,
Environment.ProcessorCount * 2, Environment.ProcessorCount);
return msa.AlignedSequences;
}
/// <summary>
/// Get the aligned sequence for stage1
/// </summary>
/// <returns>Sequence list</returns>
private List<ISequence> GetStage1AlignedSequence()
{
// MSA aligned sequences.
var msa =
new PAMSAMMultipleSequenceAligner(lstSequences,
kmerLength, DistanceFunctionTypes.EuclideanDistance,
UpdateDistanceMethodsTypes.Average,
ProfileAlignerNames.NeedlemanWunschProfileAligner,
ProfileScoreFunctionNames.InnerProduct, similarityMatrix,
gapOpenPenalty,
gapExtendPenalty, 2, 2);
return msa.AlignedSequencesA;
}
/// <summary>
/// Get Pamsam aligned sequences
/// </summary>
/// <param name="sequences">sequences.</param>
/// <returns>returns aligned sequences</returns>
private IList<ISequence> GetPAMSAMAlignedSequences(IList<ISequence> sequences)
{
similarityMatrix = new SimilarityMatrix(
SimilarityMatrix.StandardSimilarityMatrix.AmbiguousDna);
// MSA aligned sequences.
var msa = new PAMSAMMultipleSequenceAligner(sequences,
kmerLength, DistanceFunctionTypes.EuclideanDistance,
UpdateDistanceMethodsTypes.Average,
ProfileAlignerNames.NeedlemanWunschProfileAligner,
ProfileScoreFunctionNames.InnerProductFast, similarityMatrix,
gapOpenPenalty, gapExtendPenalty, 2, 2);
return msa.AlignedSequences;
}
/// <summary>
/// Validate Stage 2 aligned sequences and score of Muscle multiple sequence alignment.
/// </summary>
/// <param name="nodeName">xml node name</param>
/// <param name="expectedScoreNode">Expected score node</param>
/// <param name="hierarchicalClusteringMethodName">hierarchical clustering method name</param>
/// <param name="distanceFunctionName">kmerdistancematrix method name.</param>
/// <param name="profileAlignerName">SW/NW profiler</param>
/// <param name="profileScoreName">Profile score function name.</param>
private void ValidatePamsamAlignStage2(string nodeName,
string expectedScoreNode,
UpdateDistanceMethodsTypes hierarchicalClusteringMethodName,
DistanceFunctionTypes distanceFunctionName,
ProfileAlignerNames profileAlignerName,
ProfileScoreFunctionNames profileScoreName)
{
Initialize(nodeName, expectedScoreNode);
InitializeStage2Variables(nodeName);
// MSA aligned sequences.
var msa =
new PAMSAMMultipleSequenceAligner(lstSequences,
kmerLength, distanceFunctionName, hierarchicalClusteringMethodName,
profileAlignerName, profileScoreName, similarityMatrix, gapOpenPenalty,
gapExtendPenalty, 2, 2);
// Validate the aligned Sequence and score of stage2
if (null != msa.AlignedSequencesB)
{
Assert.AreEqual(stage2ExpectedSequences.Count, msa.AlignedSequencesB.Count);
int index = 0;
foreach (ISequence seq in msa.AlignedSequencesB)
{
Assert.AreEqual(new string(stage2ExpectedSequences[index].Select(a => (char) a).ToArray()),
new string(seq.Select(a => (char) a).ToArray()));
index++;
}
Assert.AreEqual(stage2ExpectedScore, msa.AlignmentScoreB.ToString((IFormatProvider) null));
}
ApplicationLog.WriteLine(String.Format(null,
"PamsamBvtTest:: Pamsam stage2 alignment completed successfully with all default params"));
}
/// <summary>
/// Validate Stage 3 aligned sequences and score of Muscle multiple sequence alignment.
/// </summary>
/// <param name="nodeName">xml node name</param>
/// <param name="expectedScoreNode">Expected score node</param>
/// <param name="hierarchicalClusteringMethodName">hierarchical clustering method name</param>
/// <param name="distanceFunctionName">kmerdistancematrix method name.</param>
/// <param name="profileAlignerName">SW/NW profiler</param>
/// <param name="profileScoreName">Profile score function name.</param>
private void ValidatePamsamAlignStage3(string nodeName,
string expectedScoreNode,
UpdateDistanceMethodsTypes hierarchicalClusteringMethodName,
DistanceFunctionTypes distanceFunctionName,
ProfileAlignerNames profileAlignerName,
ProfileScoreFunctionNames profileScoreName)
{
Initialize(nodeName, expectedScoreNode);
// MSA aligned sequences.
var msa =
new PAMSAMMultipleSequenceAligner(lstSequences,
kmerLength, distanceFunctionName, hierarchicalClusteringMethodName,
profileAlignerName, profileScoreName, similarityMatrix, gapOpenPenalty,
gapExtendPenalty, 2, 2);
string expectedSeqString = expectedSequences.Aggregate(string.Empty,
(current, seq) =>
current +
(new string(seq.Select(a => (char) a).ToArray()) +
","));
foreach (ISequence seq in msa.AlignedSequencesC)
{
Assert.IsTrue(expectedSeqString.Contains(new string(seq.Select(a => (char) a).ToArray())));
}
Assert.IsTrue(expectedScore.Contains(msa.AlignmentScoreC.ToString((IFormatProvider) null)));
ApplicationLog.WriteLine(String.Format(null,
"PamsamBvtTest:: Pamsam stage3 alignment completed successfully with all default params"));
}
/// <summary>
/// Validate Muscle multiple sequence alignment.
/// </summary>
/// <param name="nodeName">xml node name</param>
/// <param name="expectedScoreNode">Expected score node</param>
/// <param name="hierarchicalClusteringMethodName">hierarchical clustering method name</param>
/// <param name="distanceFunctionName">kmerdistancematrix method name.</param>
/// <param name="profileAlignerName">SW/NW profiler</param>
/// <param name="profileScoreName">Profile score function name.</param>
/// <param name="isWeightedProduct">True if it of the WeightedProduct type else false.</param>
private void ValidatePamsamAlign(string nodeName,
string expectedScoreNode,
UpdateDistanceMethodsTypes hierarchicalClusteringMethodName,
DistanceFunctionTypes distanceFunctionName,
ProfileAlignerNames profileAlignerName,
ProfileScoreFunctionNames profileScoreName,
bool isWeightedProduct)
{
Initialize(nodeName, expectedScoreNode);
// MSA aligned sequences.
var msa = new PAMSAMMultipleSequenceAligner(lstSequences,
kmerLength, distanceFunctionName,
hierarchicalClusteringMethodName,
profileAlignerName, profileScoreName,
similarityMatrix, gapOpenPenalty, gapExtendPenalty, 2, 2);
int index = 0;
foreach (ISequence seq in msa.AlignedSequences)
{
if (isWeightedProduct)
{
Assert.AreEqual(new string(seq.Select(a => (char) a).ToArray()),
new string(expectedSequences[index].Select(a => (char) a).ToArray()));
index++;
}
}
Assert.IsTrue(expectedScore.Contains(msa.AlignmentScore.ToString((IFormatProvider) null)));
}
/// <summary>
/// Read from xml config and initialize all member variables
/// </summary>
/// <param name="nodeName">xml node name</param>
/// <param name="expectedScoreNode">Expected score node</param>
private void Initialize(string nodeName, string expectedScoreNode)
{
// Read all the input sequences from xml config file
IAlphabet alphabet = Utility.GetAlphabet(utilityObj.xmlUtil.GetTextValue(nodeName,
Constants.AlphabetNameNode));
string sequenceString1 = utilityObj.xmlUtil.GetTextValue(nodeName, Constants.Sequence1);
string sequenceString2 = utilityObj.xmlUtil.GetTextValue(nodeName, Constants.Sequence2);
string sequenceString3 = utilityObj.xmlUtil.GetTextValue(nodeName, Constants.Sequence3);
string sequenceString4 = utilityObj.xmlUtil.GetTextValue(nodeName, Constants.Sequence4);
string sequenceString5 = utilityObj.xmlUtil.GetTextValue(nodeName, Constants.Sequence5);
string sequenceString6 = utilityObj.xmlUtil.GetTextValue(nodeName, Constants.Sequence6);
string sequenceString7 = utilityObj.xmlUtil.GetTextValue(nodeName, Constants.Sequence7);
string sequenceString8 = null;
string sequenceString9 = null;
// Get all the input sequence object
lstSequences = new List<ISequence>();
ISequence seq1 = new Sequence(alphabet, sequenceString1);
ISequence seq2 = new Sequence(alphabet, sequenceString2);
ISequence seq3 = new Sequence(alphabet, sequenceString3);
ISequence seq4 = new Sequence(alphabet, sequenceString4);
ISequence seq5 = new Sequence(alphabet, sequenceString5);
ISequence seq6 = new Sequence(alphabet, sequenceString6);
ISequence seq7 = new Sequence(alphabet, sequenceString7);
ISequence seq8 = null;
ISequence seq9 = null;
// Add all sequences to list.
lstSequences.Add(seq1);
lstSequences.Add(seq2);
lstSequences.Add(seq3);
lstSequences.Add(seq4);
lstSequences.Add(seq5);
lstSequences.Add(seq6);
lstSequences.Add(seq7);
similarityMatrix = new SimilarityMatrix(SimilarityMatrix.StandardSimilarityMatrix.AmbiguousDna);
profileAligner = new NeedlemanWunschProfileAlignerParallel(similarityMatrix,
ProfileScoreFunctionNames.InnerProduct,
gapOpenPenalty, gapExtendPenalty,
Environment.ProcessorCount);
// Read all expected Sequences
sequenceString1 = utilityObj.xmlUtil.GetTextValue(nodeName, Constants.ExpectedSequenceNode1);
sequenceString2 = utilityObj.xmlUtil.GetTextValue(nodeName, Constants.ExpectedSequenceNode2);
sequenceString3 = utilityObj.xmlUtil.GetTextValue(nodeName, Constants.ExpectedSequenceNode3);
sequenceString4 = utilityObj.xmlUtil.GetTextValue(nodeName, Constants.ExpectedSequenceNode4);
sequenceString5 = utilityObj.xmlUtil.GetTextValue(nodeName, Constants.ExpectedSequenceNode5);
sequenceString6 = utilityObj.xmlUtil.GetTextValue(nodeName, Constants.ExpectedSequenceNode6);
sequenceString7 = utilityObj.xmlUtil.GetTextValue(nodeName, Constants.ExpectedSequenceNode7);
sequenceString8 = utilityObj.xmlUtil.GetTextValue(nodeName, Constants.ExpectedSequenceNode8);
sequenceString9 = utilityObj.xmlUtil.GetTextValue(nodeName, Constants.ExpectedSequenceNode9);
seq1 = new Sequence(alphabet, sequenceString1);
seq2 = new Sequence(alphabet, sequenceString2);
seq3 = new Sequence(alphabet, sequenceString3);
seq4 = new Sequence(alphabet, sequenceString4);
seq5 = new Sequence(alphabet, sequenceString5);
seq6 = new Sequence(alphabet, sequenceString6);
seq7 = new Sequence(alphabet, sequenceString7);
seq8 = new Sequence(alphabet, sequenceString8);
seq9 = new Sequence(alphabet, sequenceString9);
// Add all sequences to list.
expectedSequences = new List<ISequence>();
expectedSequences.Add(seq1);
expectedSequences.Add(seq2);
expectedSequences.Add(seq3);
expectedSequences.Add(seq4);
expectedSequences.Add(seq5);
expectedSequences.Add(seq6);
expectedSequences.Add(seq7);
expectedSequences.Add(seq8);
expectedSequences.Add(seq9);
expectedScore = utilityObj.xmlUtil.GetTextValue(nodeName, expectedScoreNode);
// Parallel Option will only get set if the PAMSAMMultipleSequenceAligner is getting called
// To test separately distance matrix, binary tree etc..
// Set the parallel option using below ctor.
var msa = new PAMSAMMultipleSequenceAligner(lstSequences,
kmerLength, DistanceFunctionTypes.EuclideanDistance,
UpdateDistanceMethodsTypes.Average,
ProfileAlignerNames.NeedlemanWunschProfileAligner,
ProfileScoreFunctionNames.InnerProduct, similarityMatrix,
gapOpenPenalty, gapExtendPenalty, 2, 2);
ApplicationLog.WriteLine(String.Format(null,
"Initialization of all variables successfully completed for xml node {0}",
nodeName));
}
/// <summary>
/// Get Pamsam aligned sequences
/// </summary>
/// <param name="moleculeType">Molecule Type.</param>
/// <param name="sequences">sequences.</param>
/// <returns>returns aligned sequences</returns>
private IList<ISequence> GetPAMSAMAlignedSequences(MoleculeType moleculeType,
IList<ISequence> sequences)
{
switch (moleculeType)
{
case MoleculeType.DNA:
similarityMatrix = new SimilarityMatrix(
SimilarityMatrix.StandardSimilarityMatrix.AmbiguousDna);
break;
case MoleculeType.RNA:
similarityMatrix = new SimilarityMatrix(
SimilarityMatrix.StandardSimilarityMatrix.AmbiguousRna);
break;
case MoleculeType.Protein:
similarityMatrix = new SimilarityMatrix(
SimilarityMatrix.StandardSimilarityMatrix.Blosum62);
break;
}
// MSA aligned sequences.
var msa = new PAMSAMMultipleSequenceAligner(sequences,
kmerLength, DistanceFunctionTypes.EuclideanDistance,
UpdateDistanceMethodsTypes.Average,
ProfileAlignerNames.NeedlemanWunschProfileAligner,
ProfileScoreFunctionNames.InnerProductFast, similarityMatrix,
gapOpenPenalty, gapExtendPenalty, 2, 2);
return msa.AlignedSequences;
}
/// <summary>
/// Validate Muscle multiple sequence alignment with gap open cost and penalty.
/// </summary>
/// <param name="nodeName">xml node name</param>
/// <param name="moleculeType">molecule type</param>
/// <param name="expectedScoreNode">Expected score node</param>
/// <param name="hierarchicalClusteringMethodName">hierarchical clustering method name</param>
/// <param name="distanceFunctionName">kmerdistancematrix method name.</param>
/// <param name="profileAlignerName">SW/NW profiler</param>
/// <param name="profileScoreName">Profile score function name.</param>
/// <param name="gpOpenPenalty">Gap open penalty</param>
/// <param name="gpExtendPenalty">Gap extended penalty</param>
/// <param name="IsAlignedLargeSeq">True for large sequence else false</param>
private void ValidatePamsamAlignWithGapCost(
string nodeName, MoleculeType moleculeType, string expectedScoreNode,
UpdateDistanceMethodsTypes hierarchicalClusteringMethodName,
DistanceFunctionTypes distanceFunctionName,
ProfileAlignerNames profileAlignerName,
ProfileScoreFunctionNames profileScoreName,
int gpOpenPenalty, int gpExtendPenalty, bool IsAlignedLargeSeq)
{
Initialize(nodeName, expectedScoreNode);
// MSA aligned sequences with sepcified gap costs.
var msa = new PAMSAMMultipleSequenceAligner(lstSequences,
kmerLength, distanceFunctionName,
hierarchicalClusteringMethodName,
profileAlignerName, profileScoreName, similarityMatrix,
gpOpenPenalty,
gpExtendPenalty, 2, 2);
// Validate the aligned Sequence and score
int index = 0;
foreach (ISequence seq in msa.AlignedSequences)
{
if (IsAlignedLargeSeq)
{
Assert.AreEqual(new string(seq.Select(a => (char) a).ToArray()),
new string(expectedSequences[index].Select(a => (char) a).ToArray()));
index++;
}
}
Assert.IsTrue(expectedScore.Contains(msa.AlignmentScore.ToString((IFormatProvider) null)));
ApplicationLog.WriteLine(String.Format(null,
"PamsamP1Test:: Pamsam alignment completed successfully with equal gap cost for {0} moleculetype with all default params",
moleculeType.ToString()));
}
private void ValidatePamsamAlign(
string nodeName, MoleculeType moleculeType, string expectedScoreNode,
UpdateDistanceMethodsTypes hierarchicalClusteringMethodName,
DistanceFunctionTypes distanceFunctionName,
ProfileAlignerNames profileAlignerName,
ProfileScoreFunctionNames profileScoreName, int kmrlength,
bool addOnelineSequences, bool IsAlignForMoreSeq)
{
Initialize(nodeName, expectedScoreNode);
if (addOnelineSequences)
{
AddOneLineSequences(nodeName);
}
// MSA aligned sequences.
var msa = new PAMSAMMultipleSequenceAligner(lstSequences,
kmrlength, distanceFunctionName,
hierarchicalClusteringMethodName,
profileAlignerName, profileScoreName, similarityMatrix,
gapOpenPenalty,
gapExtendPenalty, 2, 2);
// Validate the aligned Sequence and score
int index = 0;
foreach (ISequence seq in msa.AlignedSequences)
{
if (IsAlignForMoreSeq)
{
Assert.IsTrue(expectedSequences.Contains(seq));
index++;
}
}
Assert.IsTrue(expectedScore.Contains(msa.AlignmentScore.ToString((IFormatProvider) null)));
}
