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.DeleteableMsa;
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";
63 private DeleteableMsa _msa;
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 DeleteableMsa 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( _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 s = _msa.getNumberOfSequences();
176 final int x = ForesterUtil.roundToInt( s / 20.0 );
177 while ( _msa.getNumberOfSequences() > x ) {
178 final String id = to_remove_ids.get( i );
179 //~_msa = MsaMethods.removeSequence( _msa, id );
180 _msa.deleteRow( id );
181 if ( ( s < 500 ) || ( ( step > 0 ) && ( ( ( i + 1 ) % step ) == 0 ) ) ) {
183 if ( realign && ( ( step > 0 ) && ( ( ( i + 1 ) % step ) == 0 ) ) ) {
185 msa_props.add( new MsaProperties( _msa ) );
190 System.out.print( "(realigned)" );
194 msa_props.add( new MsaProperties( _msa ) );
200 System.out.println();
208 private Phylogeny inferNJphylogeny( final PWD_DISTANCE_METHOD pwd_distance_method,
210 final boolean write_matrix,
211 final String matrix_name ) {
212 BasicSymmetricalDistanceMatrix m = null;
213 switch ( pwd_distance_method ) {
214 case KIMURA_DISTANCE:
215 m = PairwiseDistanceCalculator.calcKimuraDistances( msa );
217 case POISSON_DISTANCE:
218 m = PairwiseDistanceCalculator.calcPoissonDistances( msa );
220 case FRACTIONAL_DISSIMILARITY:
221 m = PairwiseDistanceCalculator.calcFractionalDissimilarities( msa );
224 throw new IllegalArgumentException( "invalid pwd method" );
226 if ( write_matrix ) {
228 m.write( ForesterUtil.createBufferedWriter( matrix_name ) );
230 catch ( final IOException e ) {
234 final NeighborJoiningF nj = NeighborJoiningF.createInstance( false, 5 );
235 final Phylogeny phy = nj.execute( m );
239 private StringBuilder msaStatsAsSB() {
240 final StringBuilder sb = new StringBuilder();
241 sb.append( _msa.getNumberOfSequences() );
243 sb.append( _msa.getLength() );
245 sb.append( NF_4.format( MsaMethods.calcGapRatio( _msa ) ) );
247 sb.append( NF_4.format( MsaMethods.calculateIdentityRatio( 0, _msa.getLength() - 1, _msa ).arithmeticMean() ) );
251 private final void printMsaStats( final String id ) {
252 System.out.print( ForesterUtil.pad( id, 20, ' ', false ) );
253 System.out.print( "\t" );
254 final StringBuilder sb = msaStatsAsSB();
255 System.out.print( sb );
256 System.out.print( "\t" );
259 final private void printMsaStatsWriteOutfileAndRealign( final boolean realign,
260 final boolean verbose,
261 final String id ) throws IOException, InterruptedException {
268 final String s = writeOutfile();
270 System.out.print( "-> " + s + ( realign ? "\t(realigned)" : "" ) );
274 final private void realignWithMafft() throws IOException, InterruptedException {
275 // final MsaInferrer mafft = Mafft
276 // .createInstance( "/home/czmasek/SOFTWARE/MSA/MAFFT/mafft-7.130-without-extensions/scripts/mafft" );
277 final MsaInferrer mafft = Mafft.createInstance( _path_to_mafft );
278 final List<String> opts = new ArrayList<String>();
279 for( final String o : _maffts_opts.split( "\\s" ) ) {
282 _msa = DeleteableMsa.createInstance( mafft.infer( _msa.asSequenceList(), opts ) );
285 final private void removeGapColumns() {
286 _msa.deleteGapOnlyColumns();
289 final private void removeViaGapAverage( final double mean_gapiness,
291 final boolean realign,
293 final boolean verbose ) throws IOException, InterruptedException {
294 final GapContribution stats[] = calcGapContribtionsStats( norm );
295 final List<String> to_remove_ids = new ArrayList<String>();
296 for( final GapContribution gap_gontribution : stats ) {
297 to_remove_ids.add( gap_gontribution.getId() );
303 while ( MsaMethods.calcGapRatio( _msa ) > mean_gapiness ) {
304 final String id = to_remove_ids.get( i );
305 //`_msa = MsaMethods.removeSequence( _msa, id );
306 _msa.deleteRow( id );
308 if ( ( ( step > 0 ) && ( ( ( i + 1 ) % step ) == 0 ) )
309 || ( MsaMethods.calcGapRatio( _msa ) <= mean_gapiness ) ) {
310 printMsaStatsWriteOutfileAndRealign( realign, verbose, id );
312 else if ( verbose ) {
316 System.out.println();
322 final private void removeViaLength( final int length,
324 final boolean realign,
326 final boolean verbose ) throws IOException, InterruptedException {
327 final GapContribution stats[] = calcGapContribtionsStats( norm );
328 final List<String> to_remove_ids = new ArrayList<String>();
329 for( final GapContribution gap_gontribution : stats ) {
330 to_remove_ids.add( gap_gontribution.getId() );
336 while ( _msa.getLength() > length ) {
337 final String id = to_remove_ids.get( i );
338 //~_msa = MsaMethods.removeSequence( _msa, id );
339 _msa.deleteRow( id );
341 if ( ( ( step > 0 ) && ( ( ( i + 1 ) % step ) == 0 ) ) || ( _msa.getLength() <= length ) ) {
342 printMsaStatsWriteOutfileAndRealign( realign, verbose, id );
344 else if ( verbose ) {
348 System.out.println();
354 final private void removeWorstOffenders( final int to_remove,
356 final boolean realign,
358 final boolean verbose ) throws IOException, InterruptedException {
359 final GapContribution stats[] = calcGapContribtionsStats( norm );
360 final List<String> to_remove_ids = new ArrayList<String>();
361 for( int j = 0; j < to_remove; ++j ) {
362 to_remove_ids.add( stats[ j ].getId() );
363 _removed_seq_ids.add( stats[ j ].getId() );
368 for( int i = 0; i < to_remove_ids.size(); ++i ) {
369 final String id = to_remove_ids.get( i );
370 //~ _msa = MsaMethods.removeSequence( _msa, id );
371 _msa.deleteRow( id );
373 if ( ( ( step > 0 ) && ( ( ( i + 1 ) % step ) == 0 ) ) || ( i == ( to_remove_ids.size() - 1 ) ) ) {
374 printMsaStatsWriteOutfileAndRealign( realign, verbose, id );
376 else if ( verbose ) {
380 System.out.println();
385 private void setPathToMafft( final String path_to_mafft ) {
386 _path_to_mafft = path_to_mafft;
389 final private void writeMsa( final String outfile, final MSA_FORMAT format ) throws IOException {
390 final Writer w = ForesterUtil.createBufferedWriter( outfile );
391 _msa.write( w, format );
395 private String writeOutfile() throws IOException {
396 final String s = writeMsa( _out_file_base, MSA_FORMAT.PHYLIP, ".aln" );
397 //writeMsa( _out_file_base, MSA_FORMAT.FASTA, ".fasta" );
401 public final static MsaCompactor chart( final DeleteableMsa msa,
403 final boolean realign,
405 final String path_to_mafft ) throws IOException, InterruptedException {
406 final int initial_number_of_seqs = msa.getNumberOfSequences();
407 final MsaCompactor mc = new MsaCompactor( msa );
409 mc.setPathToMafft( path_to_mafft );
411 final List<MsaProperties> msa_props = mc.chart( step, realign, norm, true );
412 Chart.display( msa_props, initial_number_of_seqs );
416 // Returns null if not path found.
417 final public static String guessPathToMafft() {
419 if ( ForesterUtil.OS_NAME.toLowerCase().indexOf( "win" ) >= 0 ) {
420 path = "C:\\Program Files\\mafft-win\\mafft.bat";
421 if ( MsaInferrer.isInstalled( path ) ) {
425 path = "/home/czmasek/SOFTWARE/MSA/MAFFT/mafft-7.130-without-extensions/scripts/mafft";
426 if ( MsaInferrer.isInstalled( path ) ) {
429 path = "/usr/local/bin/mafft";
430 if ( MsaInferrer.isInstalled( path ) ) {
433 path = "/usr/bin/mafft";
434 if ( MsaInferrer.isInstalled( path ) ) {
438 if ( MsaInferrer.isInstalled( path ) ) {
442 if ( MsaInferrer.isInstalled( path ) ) {
448 public final static MsaCompactor reduceGapAverage( final DeleteableMsa msa,
449 final double max_gap_average,
451 final boolean realign,
453 final String path_to_mafft,
454 final File out ) throws IOException, InterruptedException {
455 final MsaCompactor mc = new MsaCompactor( msa );
457 mc.setPathToMafft( path_to_mafft );
459 mc.setOutFileBase( out );
460 mc.removeViaGapAverage( max_gap_average, step, realign, norm, true );
464 public final static MsaCompactor reduceLength( final DeleteableMsa msa,
467 final boolean realign,
469 final String path_to_mafft,
470 final File out ) throws IOException, InterruptedException {
471 final MsaCompactor mc = new MsaCompactor( msa );
473 mc.setPathToMafft( path_to_mafft );
475 mc.setOutFileBase( out );
476 mc.removeViaLength( length, step, realign, norm, true );
480 public final static MsaCompactor removeWorstOffenders( final DeleteableMsa msa,
481 final int worst_offenders_to_remove,
483 final boolean realign,
485 final String path_to_mafft,
486 final File out ) throws IOException, InterruptedException {
487 final MsaCompactor mc = new MsaCompactor( msa );
489 mc.setPathToMafft( path_to_mafft );
491 mc.setOutFileBase( out );
492 mc.removeWorstOffenders( worst_offenders_to_remove, step, realign, norm, true );
496 private final static void printTableHeader() {
497 System.out.print( ForesterUtil.pad( "Id", 20, ' ', false ) );
498 System.out.print( "\t" );
499 System.out.print( "Seqs" );
500 System.out.print( "\t" );
501 System.out.print( "Length" );
502 System.out.print( "\t" );
503 System.out.print( "Gaps" );
504 System.out.print( "\t" );
505 System.out.print( "MSA qual" );
506 System.out.print( "\t" );
507 System.out.println();