2 // FORESTER -- software libraries and applications
3 // for evolutionary biology research and applications.
5 // Copyright (C) 2014 Christian M. Zmasek
6 // Copyright (C) 2014 Sanford-Burnham Medical Research Institute
9 // This library is free software; you can redistribute it and/or
10 // modify it under the terms of the GNU Lesser General Public
11 // License as published by the Free Software Foundation; either
12 // version 2.1 of the License, or (at your option) any later version.
14 // This library is distributed in the hope that it will be useful,
15 // but WITHOUT ANY WARRANTY; without even the implied warranty of
16 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
17 // Lesser General Public License for more details.
19 // You should have received a copy of the GNU Lesser General Public
20 // License along with this library; if not, write to the Free Software
21 // Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA
23 // WWW: https://sites.google.com/site/cmzmasek/home/software/forester
25 package org.forester.msa_compactor;
28 import java.io.IOException;
29 import java.io.Writer;
30 import java.math.RoundingMode;
31 import java.text.DecimalFormat;
32 import java.text.NumberFormat;
33 import java.util.ArrayList;
34 import java.util.Arrays;
35 import java.util.List;
36 import java.util.SortedSet;
37 import java.util.TreeSet;
39 import org.forester.evoinference.distance.NeighborJoiningF;
40 import org.forester.evoinference.distance.PairwiseDistanceCalculator;
41 import org.forester.evoinference.distance.PairwiseDistanceCalculator.PWD_DISTANCE_METHOD;
42 import org.forester.evoinference.matrix.distance.BasicSymmetricalDistanceMatrix;
43 import org.forester.evoinference.tools.BootstrapResampler;
44 import org.forester.msa.BasicMsa;
45 import org.forester.msa.Mafft;
46 import org.forester.msa.Msa;
47 import org.forester.msa.Msa.MSA_FORMAT;
48 import org.forester.msa.MsaInferrer;
49 import org.forester.msa.MsaMethods;
50 import org.forester.msa.ResampleableMsa;
51 import org.forester.phylogeny.Phylogeny;
52 import org.forester.phylogeny.PhylogenyMethods;
53 import org.forester.sequence.Sequence;
54 import org.forester.tools.ConfidenceAssessor;
55 import org.forester.util.ForesterUtil;
57 public class MsaCompactor {
59 final private static NumberFormat NF_3 = new DecimalFormat( "#.###" );
60 final private static NumberFormat NF_4 = new DecimalFormat( "#.####" );
61 // private final String _maffts_opts = "--retree 1";
62 private final String _maffts_opts = "--auto";
64 private File _out_file_base;
65 private String _path_to_mafft;
66 private final SortedSet<String> _removed_seq_ids;
68 NF_4.setRoundingMode( RoundingMode.HALF_UP );
69 NF_3.setRoundingMode( RoundingMode.HALF_UP );
72 private MsaCompactor( final Msa msa ) {
74 _removed_seq_ids = new TreeSet<String>();
77 final public Msa getMsa() {
81 final public SortedSet<String> getRemovedSeqIds() {
82 return _removed_seq_ids;
85 final public void setOutFileBase( final File out_file_base ) {
86 _out_file_base = out_file_base;
89 final public String writeMsa( final File outfile, final MSA_FORMAT format, final String suffix ) throws IOException {
90 final Double gr = MsaMethods.calcGapRatio( _msa );
91 final String s = outfile + "_" + _msa.getNumberOfSequences() + "_" + _msa.getLength() + "_"
92 + ForesterUtil.roundToInt( gr * 100 );
93 writeMsa( s + suffix, format );
97 final int calcNonGapResidues( final Sequence seq ) {
99 for( int i = 0; i < seq.getLength(); ++i ) {
100 if ( !seq.isGapAt( i ) ) {
107 Phylogeny pi( final String matrix ) {
108 final Phylogeny master_phy = inferNJphylogeny( PWD_DISTANCE_METHOD.KIMURA_DISTANCE, _msa, true, matrix );
111 final ResampleableMsa resampleable_msa = new ResampleableMsa( ( BasicMsa ) _msa );
112 final int[][] resampled_column_positions = BootstrapResampler.createResampledColumnPositions( _msa.getLength(),
115 final Phylogeny[] eval_phys = new Phylogeny[ n ];
116 for( int i = 0; i < n; ++i ) {
117 resampleable_msa.resample( resampled_column_positions[ i ] );
118 eval_phys[ i ] = inferNJphylogeny( PWD_DISTANCE_METHOD.KIMURA_DISTANCE, resampleable_msa, false, null );
120 ConfidenceAssessor.evaluate( "bootstrap", eval_phys, master_phy, true, 1 );
121 PhylogenyMethods.extractFastaInformation( master_phy );
125 private final GapContribution[] calcGapContribtions( final boolean normalize_for_effective_seq_length ) {
126 final double gappiness[] = calcGappiness();
127 final GapContribution stats[] = new GapContribution[ _msa.getNumberOfSequences() ];
128 for( int row = 0; row < _msa.getNumberOfSequences(); ++row ) {
129 stats[ row ] = new GapContribution( _msa.getIdentifier( row ) );
130 for( int col = 0; col < _msa.getLength(); ++col ) {
131 if ( !_msa.isGapAt( row, col ) ) {
132 stats[ row ].addToValue( gappiness[ col ] );
135 if ( normalize_for_effective_seq_length ) {
136 stats[ row ].divideValue( calcNonGapResidues( _msa.getSequence( row ) ) );
139 stats[ row ].divideValue( _msa.getLength() );
145 final private GapContribution[] calcGapContribtionsStats( final boolean norm ) {
146 final GapContribution stats[] = calcGapContribtions( norm );
147 Arrays.sort( stats );
151 private final double[] calcGappiness() {
152 final int l = _msa.getLength();
153 final double gappiness[] = new double[ l ];
154 final int seqs = _msa.getNumberOfSequences();
155 for( int i = 0; i < l; ++i ) {
156 gappiness[ i ] = ( double ) MsaMethods.calcGapSumPerColumn( _msa, i ) / seqs;
161 final private List<MsaProperties> chart( final int step,
162 final boolean realign,
164 final boolean verbose ) throws IOException, InterruptedException {
165 final GapContribution stats[] = calcGapContribtionsStats( norm );
166 final List<String> to_remove_ids = new ArrayList<String>();
167 final List<MsaProperties> msa_props = new ArrayList<MsaProperties>();
168 for( final GapContribution gap_gontribution : stats ) {
169 to_remove_ids.add( gap_gontribution.getId() );
175 final int x = ForesterUtil.roundToInt( _msa.getNumberOfSequences() / 20.0 );
176 while ( _msa.getNumberOfSequences() > x ) {
177 final String id = to_remove_ids.get( i );
178 _msa = MsaMethods.removeSequence( _msa, id );
180 msa_props.add( new MsaProperties( _msa ) );
184 if ( ( step > 0 ) && ( ( ( i + 1 ) % step ) == 0 ) ) {
190 System.out.println();
197 private Phylogeny inferNJphylogeny( final PWD_DISTANCE_METHOD pwd_distance_method,
199 final boolean write_matrix,
200 final String matrix_name ) {
201 BasicSymmetricalDistanceMatrix m = null;
202 switch ( pwd_distance_method ) {
203 case KIMURA_DISTANCE:
204 m = PairwiseDistanceCalculator.calcKimuraDistances( msa );
206 case POISSON_DISTANCE:
207 m = PairwiseDistanceCalculator.calcPoissonDistances( msa );
209 case FRACTIONAL_DISSIMILARITY:
210 m = PairwiseDistanceCalculator.calcFractionalDissimilarities( msa );
213 throw new IllegalArgumentException( "invalid pwd method" );
215 if ( write_matrix ) {
217 m.write( ForesterUtil.createBufferedWriter( matrix_name ) );
219 catch ( final IOException e ) {
223 final NeighborJoiningF nj = NeighborJoiningF.createInstance( false, 5 );
224 final Phylogeny phy = nj.execute( m );
228 private StringBuilder msaStatsAsSB() {
229 final StringBuilder sb = new StringBuilder();
230 sb.append( _msa.getNumberOfSequences() );
232 sb.append( _msa.getLength() );
234 sb.append( NF_4.format( MsaMethods.calcGapRatio( _msa ) ) );
236 sb.append( NF_4.format( MsaMethods.calculateIdentityRatio( 0, _msa.getLength() - 1, _msa ).arithmeticMean() ) );
240 private final void printMsaStats( final String id ) {
241 System.out.print( ForesterUtil.pad( id, 20, ' ', false ) );
242 System.out.print( "\t" );
243 final StringBuilder sb = msaStatsAsSB();
244 System.out.print( sb );
245 System.out.print( "\t" );
248 final private void printMsaStatsWriteOutfileAndRealign( final boolean realign,
249 final boolean verbose,
250 final String id ) throws IOException, InterruptedException {
257 final String s = writeOutfile();
259 System.out.print( "-> " + s + ( realign ? " (realigned)" : "" ) );
263 final private void realignWithMafft() throws IOException, InterruptedException {
264 // final MsaInferrer mafft = Mafft
265 // .createInstance( "/home/czmasek/SOFTWARE/MSA/MAFFT/mafft-7.130-without-extensions/scripts/mafft" );
266 final MsaInferrer mafft = Mafft.createInstance( _path_to_mafft );
267 final List<String> opts = new ArrayList<String>();
268 for( final String o : _maffts_opts.split( "\\s" ) ) {
271 //opts.add( "--maxiterate" );
272 //opts.add( "1000" );
273 //opts.add( "--localpair" );
274 //opts.add( "--quiet" );
275 _msa = mafft.infer( _msa.asSequenceList(), opts );
278 final private void removeGapColumns() {
279 _msa = MsaMethods.createInstance().removeGapColumns( 1, 0, _msa );
282 final private void removeViaGapAverage( final double mean_gapiness,
284 final boolean realign,
286 final boolean verbose ) throws IOException, InterruptedException {
287 final GapContribution stats[] = calcGapContribtionsStats( norm );
288 final List<String> to_remove_ids = new ArrayList<String>();
289 for( final GapContribution gap_gontribution : stats ) {
290 to_remove_ids.add( gap_gontribution.getId() );
296 while ( MsaMethods.calcGapRatio( _msa ) > mean_gapiness ) {
297 final String id = to_remove_ids.get( i );
298 _msa = MsaMethods.removeSequence( _msa, id );
300 if ( ( ( step > 0 ) && ( ( ( i + 1 ) % step ) == 0 ) )
301 || ( MsaMethods.calcGapRatio( _msa ) <= mean_gapiness ) ) {
302 printMsaStatsWriteOutfileAndRealign( realign, verbose, id );
304 else if ( verbose ) {
308 System.out.println();
314 final private void removeViaLength( final int length,
316 final boolean realign,
318 final boolean verbose ) throws IOException, InterruptedException {
319 final GapContribution stats[] = calcGapContribtionsStats( norm );
320 final List<String> to_remove_ids = new ArrayList<String>();
321 for( final GapContribution gap_gontribution : stats ) {
322 to_remove_ids.add( gap_gontribution.getId() );
328 while ( _msa.getLength() > length ) {
329 final String id = to_remove_ids.get( i );
330 _msa = MsaMethods.removeSequence( _msa, id );
332 if ( ( ( step > 0 ) && ( ( ( i + 1 ) % step ) == 0 ) ) || ( _msa.getLength() <= length ) ) {
333 printMsaStatsWriteOutfileAndRealign( realign, verbose, id );
335 else if ( verbose ) {
339 System.out.println();
345 final private void removeWorstOffenders( final int to_remove,
347 final boolean realign,
349 final boolean verbose ) throws IOException, InterruptedException {
350 final GapContribution stats[] = calcGapContribtionsStats( norm );
351 final List<String> to_remove_ids = new ArrayList<String>();
352 for( int j = 0; j < to_remove; ++j ) {
353 to_remove_ids.add( stats[ j ].getId() );
354 _removed_seq_ids.add( stats[ j ].getId() );
359 for( int i = 0; i < to_remove_ids.size(); ++i ) {
360 final String id = to_remove_ids.get( i );
361 _msa = MsaMethods.removeSequence( _msa, id );
363 if ( ( ( step > 0 ) && ( ( ( i + 1 ) % step ) == 0 ) ) || ( i == ( to_remove_ids.size() - 1 ) ) ) {
364 printMsaStatsWriteOutfileAndRealign( realign, verbose, id );
366 else if ( verbose ) {
370 System.out.println();
375 private void setPathToMafft( final String path_to_mafft ) {
376 _path_to_mafft = path_to_mafft;
379 final private void writeMsa( final String outfile, final MSA_FORMAT format ) throws IOException {
380 final Writer w = ForesterUtil.createBufferedWriter( outfile );
381 _msa.write( w, format );
385 private String writeOutfile() throws IOException {
386 final String s = writeMsa( _out_file_base, MSA_FORMAT.PHYLIP, ".aln" );
387 //writeMsa( _out_file_base, MSA_FORMAT.FASTA, ".fasta" );
391 public final static MsaCompactor chart( final Msa msa,
393 final boolean realign,
395 final String path_to_mafft ) throws IOException, InterruptedException {
396 final MsaCompactor mc = new MsaCompactor( msa );
398 mc.setPathToMafft( path_to_mafft );
400 final List<MsaProperties> msa_props = mc.chart( step, realign, norm, true );
401 Chart.display( msa_props );
405 // Returns null if not path found.
406 final public static String guessPathToMafft() {
408 if ( ForesterUtil.OS_NAME.toLowerCase().indexOf( "win" ) >= 0 ) {
409 path = "C:\\Program Files\\mafft-win\\mafft.bat";
410 if ( MsaInferrer.isInstalled( path ) ) {
414 path = "/home/czmasek/SOFTWARE/MSA/MAFFT/mafft-7.130-without-extensions/scripts/mafft";
415 if ( MsaInferrer.isInstalled( path ) ) {
418 path = "/usr/local/bin/mafft";
419 if ( MsaInferrer.isInstalled( path ) ) {
422 path = "/usr/bin/mafft";
423 if ( MsaInferrer.isInstalled( path ) ) {
427 if ( MsaInferrer.isInstalled( path ) ) {
431 if ( MsaInferrer.isInstalled( path ) ) {
437 public final static MsaCompactor reduceGapAverage( final Msa msa,
438 final double max_gap_average,
440 final boolean realign,
442 final String path_to_mafft,
443 final File out ) throws IOException, InterruptedException {
444 final MsaCompactor mc = new MsaCompactor( msa );
446 mc.setPathToMafft( path_to_mafft );
448 mc.setOutFileBase( out );
449 mc.removeViaGapAverage( max_gap_average, step, realign, norm, true );
453 public final static MsaCompactor reduceLength( final Msa msa,
456 final boolean realign,
458 final String path_to_mafft,
459 final File out ) throws IOException, InterruptedException {
460 final MsaCompactor mc = new MsaCompactor( msa );
462 mc.setPathToMafft( path_to_mafft );
464 mc.setOutFileBase( out );
465 mc.removeViaLength( length, step, realign, norm, true );
469 public final static MsaCompactor removeWorstOffenders( final Msa msa,
470 final int worst_offenders_to_remove,
472 final boolean realign,
474 final String path_to_mafft,
475 final File out ) throws IOException, InterruptedException {
476 final MsaCompactor mc = new MsaCompactor( msa );
478 mc.setPathToMafft( path_to_mafft );
480 mc.setOutFileBase( out );
481 mc.removeWorstOffenders( worst_offenders_to_remove, step, realign, norm, true );
485 private final static void printTableHeader() {
486 System.out.print( ForesterUtil.pad( "Id", 20, ' ', false ) );
487 System.out.print( "\t" );
488 System.out.print( "Seqs" );
489 System.out.print( "\t" );
490 System.out.print( "Length" );
491 System.out.print( "\t" );
492 System.out.print( "Gaps" );
493 System.out.print( "\t" );
494 System.out.print( "MSA qual" );
495 System.out.print( "\t" );
496 System.out.println();