public void TestProgressiveAligner()
{
ISequence templateSequence = new Sequence(Alphabets.DNA, "ATGCSWRYKMBVHDN-");
Dictionary <ISequenceItem, int> itemSet = new Dictionary <ISequenceItem, int>();
for (int i = 0; i < templateSequence.Count; ++i)
{
itemSet.Add(templateSequence[i], i);
}
Profiles.ItemSet = itemSet;
SimilarityMatrix similarityMatrix = new SimilarityMatrix(SimilarityMatrix.StandardSimilarityMatrices.AmbiguousDna);
int gapOpenPenalty = -8;
int gapExtendPenalty = -1;
int kmerLength = 3;
ISequence seqA = new Sequence(Alphabets.DNA, "GGGAAAAATCAGATT");
ISequence seqB = new Sequence(Alphabets.DNA, "GGGAATCAAAATCAG");
ISequence seqC = new Sequence(Alphabets.DNA, "GGGACAAAATCAG");
List <ISequence> sequences = new List <ISequence>();
sequences.Add(seqA);
sequences.Add(seqB);
sequences.Add(seqC);
KmerDistanceMatrixGenerator kmerDistanceMatrixGenerator =
new KmerDistanceMatrixGenerator(sequences, kmerLength, MoleculeType.DNA);
kmerDistanceMatrixGenerator.GenerateDistanceMatrix(sequences);
IHierarchicalClustering hierarchicalClustering = new HierarchicalClusteringSerial(kmerDistanceMatrixGenerator.DistanceMatrix);
BinaryGuideTree tree = new BinaryGuideTree(hierarchicalClustering);
IProgressiveAligner progressiveAligner = new ProgressiveAligner(ProfileAlignerNames.NeedlemanWunschProfileAligner, similarityMatrix, gapOpenPenalty, gapExtendPenalty);
progressiveAligner.Align(sequences, tree);
ISequence expectedSeqA = new Sequence(Alphabets.DNA, "GGGA---AAAATCAGATT");
ISequence expectedSeqB = new Sequence(Alphabets.DNA, "GGGAATCAAAATCAG---");
ISequence expectedSeqC = new Sequence(Alphabets.DNA, "GGGA--CAAAATCAG---");
Assert.AreEqual(expectedSeqA.ToString(), progressiveAligner.AlignedSequences[0].ToString());
Assert.AreEqual(expectedSeqB.ToString(), progressiveAligner.AlignedSequences[1].ToString());
Assert.AreEqual(expectedSeqC.ToString(), progressiveAligner.AlignedSequences[2].ToString());
}
public void TestProgressiveAligner()
{
MsaUtils.SetProfileItemSets(MoleculeType.DNA);
SimilarityMatrix similarityMatrix = new SimilarityMatrix(SimilarityMatrix.StandardSimilarityMatrix.AmbiguousDna);
int gapOpenPenalty = -8;
int gapExtendPenalty = -1;
int kmerLength = 4;
PAMSAMMultipleSequenceAligner.parallelOption = new ParallelOptions {
MaxDegreeOfParallelism = 2
};
ISequence seqA = new Sequence(Alphabets.DNA, "GGGAAAAATCAGATT");
ISequence seqB = new Sequence(Alphabets.DNA, "GGGAATCAAAATCAG");
ISequence seqC = new Sequence(Alphabets.DNA, "GGGACAAAATCAG");
List <ISequence> sequences = new List <ISequence>();
sequences.Add(seqA);
sequences.Add(seqB);
sequences.Add(seqC);
KmerDistanceMatrixGenerator kmerDistanceMatrixGenerator =
new KmerDistanceMatrixGenerator(sequences, kmerLength, MoleculeType.DNA);
kmerDistanceMatrixGenerator.GenerateDistanceMatrix(sequences);
IHierarchicalClustering hierarchicalClustering = new HierarchicalClusteringParallel(kmerDistanceMatrixGenerator.DistanceMatrix);
BinaryGuideTree tree = new BinaryGuideTree(hierarchicalClustering);
IProgressiveAligner progressiveAligner = new ProgressiveAligner(ProfileAlignerNames.NeedlemanWunschProfileAligner, similarityMatrix, gapOpenPenalty, gapExtendPenalty);
progressiveAligner.Align(sequences, tree);
ISequence expectedSeqA = new Sequence(Alphabets.DNA, "GGGA---AAAATCAGATT");
ISequence expectedSeqB = new Sequence(Alphabets.DNA, "GGGAATCAAAATCAG---");
ISequence expectedSeqC = new Sequence(Alphabets.DNA, "GGGA--CAAAATCAG---");
Assert.AreEqual(expectedSeqA.ToString(), progressiveAligner.AlignedSequences[0].ToString());
Assert.AreEqual(expectedSeqB.ToString(), progressiveAligner.AlignedSequences[1].ToString());
Assert.AreEqual(expectedSeqC.ToString(), progressiveAligner.AlignedSequences[2].ToString());
sequences = new List <ISequence>();
sequences.Add(new Sequence(Alphabets.DNA, "GGGAAAAATCAGATT"));
sequences.Add(new Sequence(Alphabets.DNA, "GGGAATCAAAATCAG"));
sequences.Add(new Sequence(Alphabets.DNA, "GGGAATCAAAATCAG"));
sequences.Add(new Sequence(Alphabets.DNA, "GGGAAATCG"));
sequences.Add(new Sequence(Alphabets.DNA, "GGGAATCAATCAG"));
sequences.Add(new Sequence(Alphabets.DNA, "GGGAATCTTATCAG"));
sequences.Add(new Sequence(Alphabets.DNA, "GGGACAAAATCAG"));
sequences.Add(new Sequence(Alphabets.DNA, "GGGAAAAATCAGATT"));
sequences.Add(new Sequence(Alphabets.DNA, "GGGAATCAAAATCAG"));
sequences.Add(new Sequence(Alphabets.DNA, "GGGACAAAATCAG"));
kmerDistanceMatrixGenerator.GenerateDistanceMatrix(sequences);
hierarchicalClustering = new HierarchicalClusteringParallel(kmerDistanceMatrixGenerator.DistanceMatrix);
tree = new BinaryGuideTree(hierarchicalClustering);
for (int i = 0; i < tree.NumberOfNodes; ++i)
{
Console.WriteLine("Node {0} ID: {1}", i, tree.Nodes[i].ID);
}
for (int i = 0; i < tree.NumberOfEdges; ++i)
{
Console.WriteLine("Edge {0} ID: {1}, length: {2}", i, tree.Edges[i].ID, tree.Edges[i].Length);
}
SequenceWeighting sw = new SequenceWeighting(tree);
for (int i = 0; i < sw.Weights.Length; ++i)
{
Console.WriteLine("weights {0} is {1}", i, sw.Weights[i]);
}
progressiveAligner = new ProgressiveAligner(ProfileAlignerNames.NeedlemanWunschProfileAligner, similarityMatrix, gapOpenPenalty, gapExtendPenalty);
progressiveAligner.Align(sequences, tree);
for (int i = 0; i < progressiveAligner.AlignedSequences.Count; ++i)
{
Console.WriteLine(progressiveAligner.AlignedSequences[i].ToString());
}
MsaUtils.SetProfileItemSets(MoleculeType.Protein);
ISequenceParser parser = new FastaParser();
string filepath = @"testdata\FASTA\Protein\BB11001.tfa";
IList <ISequence> orgSequences = parser.Parse(filepath);
sequences = MsaUtils.UnAlign(orgSequences);
similarityMatrix = new SimilarityMatrix(SimilarityMatrix.StandardSimilarityMatrix.Blosum62);
kmerLength = 4;
int numberOfSequences = orgSequences.Count;
gapOpenPenalty = -13;
gapExtendPenalty = -5;
kmerDistanceMatrixGenerator =
new KmerDistanceMatrixGenerator(sequences, kmerLength, MoleculeType.DNA);
kmerDistanceMatrixGenerator.GenerateDistanceMatrix(sequences);
hierarchicalClustering = new HierarchicalClusteringParallel(kmerDistanceMatrixGenerator.DistanceMatrix);
tree = new BinaryGuideTree(hierarchicalClustering);
for (int i = tree.NumberOfLeaves; i < tree.Nodes.Count; ++i)
{
Console.WriteLine("Node {0}: leftchildren-{1}, rightChildren-{2}", i, tree.Nodes[i].LeftChildren.ID, tree.Nodes[i].RightChildren.ID);
}
progressiveAligner = new ProgressiveAligner(ProfileAlignerNames.NeedlemanWunschProfileAligner, similarityMatrix, gapOpenPenalty, gapExtendPenalty);
progressiveAligner.Align(sequences, tree);
for (int i = 0; i < progressiveAligner.AlignedSequences.Count; ++i)
{
Console.WriteLine(progressiveAligner.AlignedSequences[i].ToString());
}
}
public void TestHierarchicalClusteringSerial()
{
int dimension = 4;
IDistanceMatrix distanceMatrix = new SymmetricDistanceMatrix(dimension);
for (int i = 0; i < distanceMatrix.Dimension - 1; ++i)
{
for (int j = i + 1; j < distanceMatrix.Dimension; ++j)
{
distanceMatrix[i, j] = i + j;
distanceMatrix[j, i] = i + j;
}
}
PAMSAMMultipleSequenceAligner.ParallelOption = new ParallelOptions {
MaxDegreeOfParallelism = 2
};
IHierarchicalClustering hierarchicalClustering = new HierarchicalClusteringParallel(distanceMatrix);
Assert.AreEqual(7, hierarchicalClustering.Nodes.Count);
for (int i = 0; i < dimension * 2 - 1; ++i)
{
Assert.AreEqual(i, hierarchicalClustering.Nodes[i].ID);
}
for (int i = dimension; i < hierarchicalClustering.Nodes.Count; ++i)
{
Console.WriteLine(hierarchicalClustering.Nodes[i].LeftChildren.ID);
Console.WriteLine(hierarchicalClustering.Nodes[i].RightChildren.ID);
}
// Test on sequences
ISequence seqA = new Sequence(Alphabets.DNA, "GGGAAAAATCAGATT");
ISequence seqB = new Sequence(Alphabets.DNA, "GGGAATCAAAATCAG");
ISequence seqC = new Sequence(Alphabets.DNA, "GGGAATCAAAATCAG");
List <ISequence> sequences = new List <ISequence>();
sequences.Add(seqA);
sequences.Add(seqB);
sequences.Add(seqC);
sequences.Add(new Sequence(Alphabets.DNA, "GGGAAATCG"));
sequences.Add(new Sequence(Alphabets.DNA, "GGGAATCAATCAG"));
sequences.Add(new Sequence(Alphabets.DNA, "GGGAATCTTATCAG"));
sequences.Add(new Sequence(Alphabets.DNA, "GGGACAAAATCAG"));
sequences.Add(new Sequence(Alphabets.DNA, "GGGAAAAATCAGATT"));
sequences.Add(new Sequence(Alphabets.DNA, "GGGAATCAAAATCAG"));
sequences.Add(new Sequence(Alphabets.DNA, "GGGACAAAATCAG"));
int kmerLength = 4;
KmerDistanceMatrixGenerator kmerDistanceMatrixGenerator =
new KmerDistanceMatrixGenerator(sequences, kmerLength, Alphabets.AmbiguousDNA);
//Console.WriteLine(kmerDistanceMatrixGenerator.Name);
kmerDistanceMatrixGenerator.GenerateDistanceMatrix(sequences);
//Console.WriteLine(kmerDistanceMatrixGenerator.DistanceMatrix);
for (int i = 0; i < kmerDistanceMatrixGenerator.DistanceMatrix.Dimension - 1; ++i)
{
for (int j = i + 1; j < kmerDistanceMatrixGenerator.DistanceMatrix.Dimension; ++j)
{
Console.WriteLine("{0}-{1}: {2}", i, j, kmerDistanceMatrixGenerator.DistanceMatrix[i, j]);
}
}
hierarchicalClustering = new HierarchicalClusteringParallel(kmerDistanceMatrixGenerator.DistanceMatrix);
for (int i = 0; i < hierarchicalClustering.Nodes.Count; ++i)
{
Assert.AreEqual(true, hierarchicalClustering.Nodes[i].NeedReAlignment);
}
BinaryGuideTree tree = new BinaryGuideTree(hierarchicalClustering);
for (int i = 0; i < tree.Nodes.Count; ++i)
{
Assert.AreEqual(true, tree.Nodes[i].NeedReAlignment);
}
// SimilarityMatrix similarityMatrix = new SimilarityMatrix(SimilarityMatrix.StandardSimilarityMatrix.AmbiguousDna);
//Assert.AreEqual(0, hierarchicalClustering.Nodes[4].LeftChildren.ID);
//Assert.AreEqual(1, hierarchicalClustering.Nodes[4].RightChildren.ID);
//Assert.AreEqual(2, hierarchicalClustering.Nodes[5].LeftChildren.ID);
//Assert.AreEqual(4, hierarchicalClustering.Nodes[5].RightChildren.ID);
//Assert.AreEqual(3, hierarchicalClustering.Nodes[6].LeftChildren.ID);
//Assert.AreEqual(5, hierarchicalClustering.Nodes[6].RightChildren.ID);
// Test on larger dataset
string filepath = @"TestUtils\Fasta\RV11_BBS_all.afa".TestDir();
FastAParser parser = new FastAParser();
IList <ISequence> orgSequences = parser.Parse(filepath).ToList();
sequences = MsaUtils.UnAlign(orgSequences);
kmerDistanceMatrixGenerator =
new KmerDistanceMatrixGenerator(sequences, kmerLength, Alphabets.AmbiguousDNA);
kmerDistanceMatrixGenerator.GenerateDistanceMatrix(sequences);
hierarchicalClustering = new HierarchicalClusteringParallel(kmerDistanceMatrixGenerator.DistanceMatrix);
for (int i = sequences.Count; i < hierarchicalClustering.Nodes.Count; ++i)
{
Console.WriteLine("Node {0}: leftchildren-{1}, rightChildren-{2}", i, hierarchicalClustering.Nodes[i].LeftChildren.ID, hierarchicalClustering.Nodes[i].RightChildren.ID);
}
}