import java.text.NumberFormat;
import java.util.ArrayList;
import java.util.Arrays;
-import java.util.Comparator;
import java.util.List;
import java.util.SortedSet;
import java.util.TreeSet;
+import org.forester.archaeopteryx.Archaeopteryx;
+import org.forester.evoinference.distance.NeighborJoiningF;
+import org.forester.evoinference.distance.PairwiseDistanceCalculator;
+import org.forester.evoinference.distance.PairwiseDistanceCalculator.PWD_DISTANCE_METHOD;
+import org.forester.evoinference.matrix.distance.BasicSymmetricalDistanceMatrix;
+import org.forester.evoinference.tools.BootstrapResampler;
+import org.forester.msa.BasicMsa;
import org.forester.msa.Mafft;
import org.forester.msa.Msa;
import org.forester.msa.Msa.MSA_FORMAT;
import org.forester.msa.MsaInferrer;
import org.forester.msa.MsaMethods;
+import org.forester.msa.ResampleableMsa;
+import org.forester.phylogeny.Phylogeny;
+import org.forester.phylogeny.PhylogenyMethods;
import org.forester.sequence.Sequence;
-import org.forester.util.BasicDescriptiveStatistics;
-import org.forester.util.DescriptiveStatistics;
+import org.forester.tools.ConfidenceAssessor;
import org.forester.util.ForesterUtil;
public class MsaCompactor {
return ng;
}
- private final DescriptiveStatistics[] calcGapContribtionsX( final boolean normalize_for_effective_seq_length ) {
- final double gappiness[] = calcGappiness();
- final DescriptiveStatistics stats[] = new DescriptiveStatistics[ _msa.getNumberOfSequences() ];
- for( int row = 0; row < _msa.getNumberOfSequences(); ++row ) {
- stats[ row ] = new BasicDescriptiveStatistics( _msa.getIdentifier( row ) );
- final double l = calculateEffectiveLengthRatio( row );
- for( int col = 0; col < _msa.getLength(); ++col ) {
- if ( !_msa.isGapAt( row, col ) ) {
- if ( normalize_for_effective_seq_length ) {
- stats[ row ].addValue( gappiness[ col ] / l );
- }
- else {
- stats[ row ].addValue( gappiness[ col ] );
- }
- }
- }
- }
- return stats;
- }
-
private final GapContribution[] calcGapContribtions( final boolean normalize_for_effective_seq_length ) {
final double gappiness[] = calcGappiness();
final GapContribution stats[] = new GapContribution[ _msa.getNumberOfSequences() ];
}
}
if ( normalize_for_effective_seq_length ) {
- stats[ row ].divideValue( calculateEffectiveLengthRatio( row ) );
+ stats[ row ].divideValue( calcNonGapResidues( _msa.getSequence( row ) ) );
}
else {
- //
+ stats[ row ].divideValue( _msa.getLength() );
}
}
return stats;
return gappiness;
}
- private double calculateEffectiveLengthRatio( final int row ) {
- return ( double ) calcNonGapResidues( _msa.getSequence( row ) ) / _msa.getLength();
- }
-
final private void mafft() throws IOException, InterruptedException {
final MsaInferrer mafft = Mafft
.createInstance( "/home/czmasek/SOFTWARE/MSA/MAFFT/mafft-7.130-without-extensions/scripts/mafft" );
final List<String> opts = new ArrayList<String>();
- // opts.add( "--maxiterate" );
- // opts.add( "1000" );
- // opts.add( "--localpair" );
+ opts.add( "--maxiterate" );
+ opts.add( "1000" );
+ opts.add( "--localpair" );
opts.add( "--quiet" );
_msa = mafft.infer( _msa.asSequenceList(), opts );
}
}
}
+ Phylogeny pi( final String matrix ) {
+ final Phylogeny master_phy = inferNJphylogeny( PWD_DISTANCE_METHOD.KIMURA_DISTANCE, _msa, true, matrix );
+ final int seed = 15;
+ final int n = 100;
+ final ResampleableMsa resampleable_msa = new ResampleableMsa( ( BasicMsa ) _msa );
+ final int[][] resampled_column_positions = BootstrapResampler.createResampledColumnPositions( _msa.getLength(),
+ n,
+ seed );
+ final Phylogeny[] eval_phys = new Phylogeny[ n ];
+ for( int i = 0; i < n; ++i ) {
+ resampleable_msa.resample( resampled_column_positions[ i ] );
+ eval_phys[ i ] = inferNJphylogeny( PWD_DISTANCE_METHOD.KIMURA_DISTANCE, resampleable_msa, false, null );
+ }
+ ConfidenceAssessor.evaluate( "bootstrap", eval_phys, master_phy, true, 1 );
+ PhylogenyMethods.extractFastaInformation( master_phy );
+ return master_phy;
+ }
+
+ private Phylogeny inferNJphylogeny( final PWD_DISTANCE_METHOD pwd_distance_method,
+ final Msa msa,
+ final boolean write_matrix,
+ final String matrix_name ) {
+ BasicSymmetricalDistanceMatrix m = null;
+ switch ( pwd_distance_method ) {
+ case KIMURA_DISTANCE:
+ m = PairwiseDistanceCalculator.calcKimuraDistances( msa );
+ break;
+ case POISSON_DISTANCE:
+ m = PairwiseDistanceCalculator.calcPoissonDistances( msa );
+ break;
+ case FRACTIONAL_DISSIMILARITY:
+ m = PairwiseDistanceCalculator.calcFractionalDissimilarities( msa );
+ break;
+ default:
+ throw new IllegalArgumentException( "invalid pwd method" );
+ }
+ if ( write_matrix ) {
+ try {
+ m.write( ForesterUtil.createBufferedWriter( matrix_name ) );
+ }
+ catch ( final IOException e ) {
+ // TODO Auto-generated catch block
+ e.printStackTrace();
+ }
+ }
+ final NeighborJoiningF nj = NeighborJoiningF.createInstance( false, 5 );
+ final Phylogeny phy = nj.execute( m );
+ return phy;
+ }
+
final private void removeWorstOffenders( final int to_remove,
final int step,
final boolean realign,
final boolean norm ) throws IOException, InterruptedException {
+ final Phylogeny a = pi( "a.pwd" );
+ Archaeopteryx.createApplication( a );
final GapContribution stats[] = calcGapContribtionsStats( norm );
final List<String> to_remove_ids = new ArrayList<String>();
for( int j = 0; j < to_remove; ++j ) {
if ( realign ) {
mafft();
}
+ final Phylogeny b = pi( "b.pwd" );
+ Archaeopteryx.createApplication( b );
}
final private void writeMsa( final String outfile, final MSA_FORMAT format ) throws IOException {
mc.removeWorstOffenders( worst_offenders_to_remove, 1, realign, norm );
return mc;
}
-
- public static enum SORT_BY {
- MAX, MEAN, MEDIAN;
- }
-
- final static class DescriptiveStatisticsComparator implements Comparator<DescriptiveStatistics> {
-
- final private boolean _ascending;
- final private SORT_BY _sort_by;
-
- public DescriptiveStatisticsComparator( final boolean ascending, final SORT_BY sort_by ) {
- _ascending = ascending;
- _sort_by = sort_by;
- }
-
- @Override
- public final int compare( final DescriptiveStatistics s0, final DescriptiveStatistics s1 ) {
- switch ( _sort_by ) {
- case MAX:
- if ( s0.getMax() < s1.getMax() ) {
- return _ascending ? -1 : 1;
- }
- else if ( s0.getMax() > s1.getMax() ) {
- return _ascending ? 1 : -1;
- }
- return 0;
- case MEAN:
- if ( s0.arithmeticMean() < s1.arithmeticMean() ) {
- return _ascending ? -1 : 1;
- }
- else if ( s0.arithmeticMean() > s1.arithmeticMean() ) {
- return _ascending ? 1 : -1;
- }
- return 0;
- case MEDIAN:
- if ( s0.median() < s1.median() ) {
- return _ascending ? -1 : 1;
- }
- else if ( s0.median() > s1.median() ) {
- return _ascending ? 1 : -1;
- }
- return 0;
- default:
- return 0;
- }
- }
- }
}