2 * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
3 * Copyright (C) $$Year-Rel$$ The Jalview Authors
5 * This file is part of Jalview.
7 * Jalview is free software: you can redistribute it and/or
8 * modify it under the terms of the GNU General Public License
9 * as published by the Free Software Foundation, either version 3
10 * of the License, or (at your option) any later version.
12 * Jalview is distributed in the hope that it will be useful, but
13 * WITHOUT ANY WARRANTY; without even the implied warranty
14 * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
15 * PURPOSE. See the GNU General Public License for more details.
17 * You should have received a copy of the GNU General Public License
18 * along with Jalview. If not, see <http://www.gnu.org/licenses/>.
19 * The Jalview Authors are detailed in the 'AUTHORS' file.
21 package jalview.datamodel;
23 import static org.testng.AssertJUnit.assertEquals;
24 import static org.testng.AssertJUnit.assertFalse;
25 import static org.testng.AssertJUnit.assertNotNull;
26 import static org.testng.AssertJUnit.assertNull;
27 import static org.testng.AssertJUnit.assertSame;
28 import static org.testng.AssertJUnit.assertTrue;
30 import jalview.analysis.AlignmentGenerator;
31 import jalview.datamodel.AlignedCodonFrame.SequenceToSequenceMapping;
32 import jalview.gui.JvOptionPane;
33 import jalview.io.DataSourceType;
34 import jalview.io.FileFormat;
35 import jalview.io.FileFormatI;
36 import jalview.io.FormatAdapter;
37 import jalview.util.Comparison;
38 import jalview.util.MapList;
40 import java.io.IOException;
41 import java.util.Arrays;
42 import java.util.Iterator;
43 import java.util.List;
45 import org.testng.Assert;
46 import org.testng.annotations.BeforeClass;
47 import org.testng.annotations.BeforeMethod;
48 import org.testng.annotations.Test;
51 * Unit tests for Alignment datamodel.
56 public class AlignmentTest
59 @BeforeClass(alwaysRun = true)
60 public void setUpJvOptionPane()
62 JvOptionPane.setInteractiveMode(false);
63 JvOptionPane.setMockResponse(JvOptionPane.CANCEL_OPTION);
67 private static final String TEST_DATA =
69 "#=GS D.melanogaster.1 AC AY119185.1/838-902\n" +
70 "#=GS D.melanogaster.2 AC AC092237.1/57223-57161\n" +
71 "#=GS D.melanogaster.3 AC AY060611.1/560-627\n" +
72 "D.melanogaster.1 G.AGCC.CU...AUGAUCGA\n" +
73 "#=GR D.melanogaster.1 SS ................((((\n" +
74 "D.melanogaster.2 C.AUUCAACU.UAUGAGGAU\n" +
75 "#=GR D.melanogaster.2 SS ................((((\n" +
76 "D.melanogaster.3 G.UGGCGCU..UAUGACGCA\n" +
77 "#=GR D.melanogaster.3 SS (.(((...(....(((((((\n" +
80 private static final String AA_SEQS_1 =
86 private static final String CDNA_SEQS_1 =
87 ">Seq1Name/100-111\n" +
88 "AC-GG--CUC-CAA-CT\n" +
89 ">Seq2Name/200-211\n" +
90 "-CG-TTA--ACG---AAGT\n";
92 private static final String CDNA_SEQS_2 =
99 private AlignmentI al;
102 * Helper method to load an alignment and ensure dataset sequences are set up.
108 * @throws IOException
110 protected AlignmentI loadAlignment(final String data, FileFormatI format)
113 AlignmentI a = new FormatAdapter().readFile(data, DataSourceType.PASTE,
120 * assert wrapper: tests all references in the given alignment are consistent
124 public static void assertAlignmentDatasetRefs(AlignmentI alignment)
126 verifyAlignmentDatasetRefs(alignment, true, null);
130 * assert wrapper: tests all references in the given alignment are consistent
134 * - prefixed to any assert failed messages
136 public static void assertAlignmentDatasetRefs(AlignmentI alignment,
139 verifyAlignmentDatasetRefs(alignment, true, message);
143 * verify sequence and dataset references are properly contained within
147 * - the alignmentI object to verify (either alignment or dataset)
149 * - when set, testng assertions are raised.
151 * - null or a string message to prepend to the assert failed
153 * @return true if alignment references were in order, otherwise false.
155 public static boolean verifyAlignmentDatasetRefs(AlignmentI alignment,
156 boolean raiseAssert, String message)
162 if (alignment == null)
166 Assert.fail(message + "Alignment for verification was null.");
170 if (alignment.getDataset() != null)
172 AlignmentI dataset = alignment.getDataset();
173 // check all alignment sequences have their dataset within the dataset
174 for (SequenceI seq : alignment.getSequences())
176 SequenceI seqds = seq.getDatasetSequence();
177 if (seqds.getDatasetSequence() != null)
182 + " Alignment contained a sequence who's dataset sequence has a second dataset reference.");
186 if (dataset.findIndex(seqds) == -1)
191 + " Alignment contained a sequence who's dataset sequence was not in the dataset.");
196 return verifyAlignmentDatasetRefs(alignment.getDataset(),
197 raiseAssert, message);
202 // verify all dataset sequences
203 for (SequenceI seqds : alignment.getSequences())
206 if (seqds.getDatasetSequence() != null)
211 + " Dataset contained a sequence with non-null dataset reference (ie not a dataset sequence!)");
215 int foundp = alignment.findIndex(seqds);
221 + " Dataset sequence array contains a reference at "
222 + dsp + " to a sequence first seen at " + foundp + " ("
223 + seqds.toString() + ")");
227 if (seqds.getDBRefs() != null)
229 for (DBRefEntry dbr : seqds.getDBRefs())
231 if (dbr.getMap() != null)
233 SequenceI seqdbrmapto = dbr.getMap().getTo();
234 if (seqdbrmapto != null)
236 if (seqdbrmapto.getDatasetSequence() != null)
241 + " DBRefEntry for sequence in alignment had map to sequence which was not a dataset sequence");
246 if (alignment.findIndex(dbr.getMap().getTo()) == -1)
251 + " DBRefEntry for sequence in alignment had map to sequence not in dataset");
260 // finally, verify codonmappings involve only dataset sequences.
261 if (alignment.getCodonFrames() != null)
263 for (AlignedCodonFrame alc : alignment.getCodonFrames())
265 for (SequenceToSequenceMapping ssm : alc.getMappings())
267 if (ssm.getFromSeq().getDatasetSequence() != null)
272 + " CodonFrame-SSM-FromSeq is not a dataset sequence");
276 if (alignment.findIndex(ssm.getFromSeq()) == -1)
282 + " CodonFrame-SSM-FromSeq is not contained in dataset");
286 if (ssm.getMapping().getTo().getDatasetSequence() != null)
291 + " CodonFrame-SSM-Mapping-ToSeq is not a dataset sequence");
295 if (alignment.findIndex(ssm.getMapping().getTo()) == -1)
301 + " CodonFrame-SSM-Mapping-ToSeq is not contained in dataset");
309 return true; // all relationships verified!
313 * call verifyAlignmentDatasetRefs with and without assertion raising enabled,
314 * to check expected pass/fail actually occurs in both conditions
320 private void assertVerifyAlignment(AlignmentI al, boolean expected,
328 Assert.assertTrue(verifyAlignmentDatasetRefs(al, true, null),
329 "Valid test alignment failed when raiseAsserts enabled:"
331 } catch (AssertionError ae)
333 ae.printStackTrace();
335 "Valid test alignment raised assertion errors when raiseAsserts enabled: "
338 // also check validation passes with asserts disabled
339 Assert.assertTrue(verifyAlignmentDatasetRefs(al, false, null),
340 "Valid test alignment tested false when raiseAsserts disabled:"
345 boolean assertRaised = false;
348 verifyAlignmentDatasetRefs(al, true, null);
349 } catch (AssertionError ae)
351 // expected behaviour
356 Assert.fail("Invalid test alignment passed when raiseAsserts enabled:"
359 // also check validation passes with asserts disabled
360 Assert.assertFalse(verifyAlignmentDatasetRefs(al, false, null),
361 "Invalid test alignment tested true when raiseAsserts disabled:"
366 @Test(groups = { "Functional" })
367 public void testVerifyAlignmentDatasetRefs()
369 SequenceI sq1 = new Sequence("sq1", "ASFDD"), sq2 = new Sequence("sq2",
372 // construct simple valid alignment dataset
373 Alignment al = new Alignment(new SequenceI[] { sq1, sq2 });
374 // expect this to pass
375 assertVerifyAlignment(al, true, "Simple valid alignment didn't verify");
377 // check test for sequence->datasetSequence validity
378 sq1.setDatasetSequence(sq2);
379 assertVerifyAlignment(al, false,
380 "didn't detect dataset sequence with a dataset sequence reference.");
382 sq1.setDatasetSequence(null);
383 assertVerifyAlignment(
386 "didn't reinstate validity after nulling dataset sequence dataset reference");
388 // now create dataset and check again
389 al.createDatasetAlignment();
390 assertNotNull(al.getDataset());
392 assertVerifyAlignment(al, true,
393 "verify failed after createDatasetAlignment");
395 // create a dbref on sq1 with a sequence ref to sq2
396 DBRefEntry dbrs1tos2 = new DBRefEntry("UNIPROT", "1", "Q111111");
397 dbrs1tos2.setMap(new Mapping(sq2.getDatasetSequence(),
398 new int[] { 1, 5 }, new int[] { 2, 6 }, 1, 1));
399 sq1.getDatasetSequence().addDBRef(dbrs1tos2);
400 assertVerifyAlignment(al, true,
401 "verify failed after addition of valid DBRefEntry/map");
402 // now create a dbref on a new sequence which maps to another sequence
403 // outside of the dataset
404 SequenceI sqout = new Sequence("sqout", "ututututucagcagcag"), sqnew = new Sequence(
406 DBRefEntry sqnewsqout = new DBRefEntry("ENAFOO", "1", "R000001");
407 sqnewsqout.setMap(new Mapping(sqout, new int[] { 1, 6 }, new int[] { 1,
409 al.getDataset().addSequence(sqnew);
411 assertVerifyAlignment(al, true,
412 "verify failed after addition of new sequence to dataset");
413 // now start checking exception conditions
414 sqnew.addDBRef(sqnewsqout);
415 assertVerifyAlignment(
418 "verify passed when a dbref with map to sequence outside of dataset was added");
419 // make the verify pass by adding the outsider back in
420 al.getDataset().addSequence(sqout);
421 assertVerifyAlignment(al, true,
422 "verify should have passed after adding dbref->to sequence in to dataset");
423 // and now the same for a codon mapping...
424 SequenceI sqanotherout = new Sequence("sqanotherout",
425 "aggtutaggcagcagcag");
427 AlignedCodonFrame alc = new AlignedCodonFrame();
428 alc.addMap(sqanotherout, sqnew, new MapList(new int[] { 1, 6 },
429 new int[] { 1, 18 }, 3, 1));
431 al.addCodonFrame(alc);
432 Assert.assertEquals(al.getDataset().getCodonFrames().size(), 1);
434 assertVerifyAlignment(
437 "verify passed when alCodonFrame mapping to sequence outside of dataset was added");
438 // make the verify pass by adding the outsider back in
439 al.getDataset().addSequence(sqanotherout);
440 assertVerifyAlignment(
443 "verify should have passed once all sequences involved in alCodonFrame were added to dataset");
444 al.getDataset().addSequence(sqanotherout);
445 assertVerifyAlignment(al, false,
446 "verify should have failed when a sequence was added twice to the dataset");
447 al.getDataset().deleteSequence(sqanotherout);
448 assertVerifyAlignment(al, true,
449 "verify should have passed after duplicate entry for sequence was removed");
453 * checks that the sequence data for an alignment's dataset is non-redundant.
454 * Fails if there are sequences with same id, sequence, start, and.
457 public static void assertDatasetIsNormalised(AlignmentI al)
459 assertDatasetIsNormalised(al, null);
463 * checks that the sequence data for an alignment's dataset is non-redundant.
464 * Fails if there are sequences with same id, sequence, start, and.
467 * - alignment to verify
469 * - null or message prepended to exception message.
471 public static void assertDatasetIsNormalised(AlignmentI al, String message)
473 if (al.getDataset() != null)
475 assertDatasetIsNormalised(al.getDataset(), message);
479 * look for pairs of sequences with same ID, start, end, and sequence
481 List<SequenceI> seqSet = al.getSequences();
482 for (int p = 0; p < seqSet.size(); p++)
484 SequenceI pSeq = seqSet.get(p);
485 for (int q = p + 1; q < seqSet.size(); q++)
487 SequenceI qSeq = seqSet.get(q);
488 if (pSeq.getStart() != qSeq.getStart())
492 if (pSeq.getEnd() != qSeq.getEnd())
496 if (!pSeq.getName().equals(qSeq.getName()))
500 if (!Arrays.equals(pSeq.getSequence(), qSeq.getSequence()))
504 Assert.fail((message == null ? "" : message + " :")
505 + "Found similar sequences at position " + p + " and " + q
506 + "\n" + pSeq.toString());
511 @Test(groups = { "Functional", "Asserts" })
512 public void testAssertDatasetIsNormalised()
514 Sequence sq1 = new Sequence("s1/1-4", "asdf");
515 Sequence sq1shift = new Sequence("s1/2-5", "asdf");
516 Sequence sq1seqd = new Sequence("s1/1-4", "asdt");
517 Sequence sq2 = new Sequence("s2/1-4", "asdf");
518 Sequence sq1dup = new Sequence("s1/1-4", "asdf");
520 Alignment al = new Alignment(new SequenceI[] { sq1 });
525 assertDatasetIsNormalised(al);
526 } catch (AssertionError ae)
528 Assert.fail("Single sequence should be valid normalised dataset.");
533 assertDatasetIsNormalised(al);
534 } catch (AssertionError ae)
536 Assert.fail("Two different sequences should be valid normalised dataset.");
539 * now change sq2's name in the alignment. should still be valid
541 al.findName(sq2.getName()).setName("sq1");
544 assertDatasetIsNormalised(al);
545 } catch (AssertionError ae)
547 Assert.fail("Two different sequences in dataset, but same name in alignment, should be valid normalised dataset.");
550 al.addSequence(sq1seqd);
553 assertDatasetIsNormalised(al);
554 } catch (AssertionError ae)
556 Assert.fail("sq1 and sq1 with different sequence should be distinct.");
559 al.addSequence(sq1shift);
562 assertDatasetIsNormalised(al);
563 } catch (AssertionError ae)
565 Assert.fail("sq1 and sq1 with different start/end should be distinct.");
568 * finally, the failure case
570 al.addSequence(sq1dup);
571 boolean ssertRaised = false;
574 assertDatasetIsNormalised(al);
576 } catch (AssertionError ae)
582 Assert.fail("Expected identical sequence to raise exception.");
587 * Read in Stockholm format test data including secondary structure
590 @BeforeMethod(alwaysRun = true)
591 public void setUp() throws IOException
593 al = loadAlignment(TEST_DATA, FileFormat.Stockholm);
595 for (AlignmentAnnotation ann : al.getAlignmentAnnotation())
597 ann.setCalcId("CalcIdFor" + al.getSequenceAt(i).getName());
603 * Test method that returns annotations that match on calcId.
605 @Test(groups = { "Functional" })
606 public void testFindAnnotation_byCalcId()
608 Iterable<AlignmentAnnotation> anns = al
609 .findAnnotation("CalcIdForD.melanogaster.2");
610 Iterator<AlignmentAnnotation> iter = anns.iterator();
611 assertTrue(iter.hasNext());
612 AlignmentAnnotation ann = iter.next();
613 assertEquals("D.melanogaster.2", ann.sequenceRef.getName());
614 assertFalse(iter.hasNext());
617 anns = al.findAnnotation("CalcIdForD.melanogaster.?");
618 assertFalse(iter.hasNext());
619 anns = al.findAnnotation(null);
620 assertFalse(iter.hasNext());
624 * Test method that returns annotations that match on reference sequence,
627 @Test(groups = { "Functional" })
628 public void testFindAnnotations_bySeqLabelandorCalcId()
630 // TODO: finish testFindAnnotations_bySeqLabelandorCalcId test
631 /* Note - this is an incomplete test - need to check null or
632 * non-null [ matches, not matches ] behaviour for each of the three
635 // search for a single, unique calcId with wildcards on other params
636 Iterable<AlignmentAnnotation> anns = al.findAnnotations(null,
637 "CalcIdForD.melanogaster.2", null);
638 Iterator<AlignmentAnnotation> iter = anns.iterator();
639 assertTrue(iter.hasNext());
640 AlignmentAnnotation ann = iter.next();
641 assertEquals("D.melanogaster.2", ann.sequenceRef.getName());
642 assertFalse(iter.hasNext());
644 // save reference to test sequence reference parameter
645 SequenceI rseq = ann.sequenceRef;
647 // search for annotation associated with a single sequence
648 anns = al.findAnnotations(rseq, null, null);
649 iter = anns.iterator();
650 assertTrue(iter.hasNext());
652 assertEquals("D.melanogaster.2", ann.sequenceRef.getName());
653 assertFalse(iter.hasNext());
655 // search for annotation with a non-existant calcId
656 anns = al.findAnnotations(null, "CalcIdForD.melanogaster.?", null);
657 iter = anns.iterator();
658 assertFalse(iter.hasNext());
660 // search for annotation with a particular label - expect three
661 anns = al.findAnnotations(null, null, "Secondary Structure");
662 iter = anns.iterator();
663 assertTrue(iter.hasNext());
665 assertTrue(iter.hasNext());
667 assertTrue(iter.hasNext());
670 assertFalse(iter.hasNext());
672 // search for annotation on one sequence with a particular label - expect
675 anns = al.findAnnotations(sqfound = al.getSequenceAt(1), null,
676 "Secondary Structure");
677 iter = anns.iterator();
678 assertTrue(iter.hasNext());
679 // expect reference to sequence 1 in the alignment
680 assertTrue(sqfound == iter.next().sequenceRef);
681 assertFalse(iter.hasNext());
683 // null on all parameters == find all annotations
684 anns = al.findAnnotations(null, null, null);
685 iter = anns.iterator();
686 int n = al.getAlignmentAnnotation().length;
687 while (iter.hasNext())
692 assertTrue("Found " + n + " fewer annotations from search.", n == 0);
695 @Test(groups = { "Functional" })
696 public void testDeleteAllAnnotations_includingAutocalculated()
698 AlignmentAnnotation aa = new AlignmentAnnotation("Consensus",
700 aa.autoCalculated = true;
701 al.addAnnotation(aa);
702 AlignmentAnnotation[] anns = al.getAlignmentAnnotation();
703 assertEquals("Wrong number of annotations before deleting", 4,
705 al.deleteAllAnnotations(true);
706 assertEquals("Not all deleted", 0, al.getAlignmentAnnotation().length);
709 @Test(groups = { "Functional" })
710 public void testDeleteAllAnnotations_excludingAutocalculated()
712 AlignmentAnnotation aa = new AlignmentAnnotation("Consensus",
714 aa.autoCalculated = true;
715 al.addAnnotation(aa);
716 AlignmentAnnotation[] anns = al.getAlignmentAnnotation();
717 assertEquals("Wrong number of annotations before deleting", 4,
719 al.deleteAllAnnotations(false);
720 assertEquals("Not just one annotation left", 1,
721 al.getAlignmentAnnotation().length);
725 * Tests for realigning as per a supplied alignment: Dna as Dna.
727 * Note: AlignedCodonFrame's state variables are named for protein-to-cDNA
728 * mapping, but can be exploited for a general 'sequence-to-sequence' mapping
731 * @throws IOException
733 @Test(groups = { "Functional" })
734 public void testAlignAs_dnaAsDna() throws IOException
737 AlignmentI al1 = loadAlignment(CDNA_SEQS_1, FileFormat.Fasta);
739 AlignmentI al2 = loadAlignment(CDNA_SEQS_2, FileFormat.Fasta);
742 * Make mappings between sequences. The 'aligned cDNA' is playing the role
743 * of what would normally be protein here.
745 makeMappings(al1, al2);
747 ((Alignment) al2).alignAs(al1, false, true);
748 assertEquals("GC-TC--GUC-GTACT", al2.getSequenceAt(0)
749 .getSequenceAsString());
750 assertEquals("-GG-GTC--AGG--CAGT", al2.getSequenceAt(1)
751 .getSequenceAsString());
755 * Aligning protein from cDNA.
757 * @throws IOException
759 @Test(groups = { "Functional" })
760 public void testAlignAs_proteinAsCdna() throws IOException
762 // see also AlignmentUtilsTests
763 AlignmentI al1 = loadAlignment(CDNA_SEQS_1, FileFormat.Fasta);
764 AlignmentI al2 = loadAlignment(AA_SEQS_1, FileFormat.Fasta);
765 makeMappings(al1, al2);
767 // Fudge - alignProteinAsCdna expects mappings to be on protein
768 al2.getCodonFrames().addAll(al1.getCodonFrames());
770 ((Alignment) al2).alignAs(al1, false, true);
771 assertEquals("K-Q-Y-L-", al2.getSequenceAt(0).getSequenceAsString());
772 assertEquals("-R-F-P-W", al2.getSequenceAt(1).getSequenceAsString());
776 * Test aligning cdna as per protein alignment.
778 * @throws IOException
780 @Test(groups = { "Functional" }, enabled = true)
781 // TODO review / update this test after redesign of alignAs method
782 public void testAlignAs_cdnaAsProtein() throws IOException
785 * Load alignments and add mappings for cDNA to protein
787 AlignmentI al1 = loadAlignment(CDNA_SEQS_1, FileFormat.Fasta);
788 AlignmentI al2 = loadAlignment(AA_SEQS_1, FileFormat.Fasta);
789 makeMappings(al1, al2);
792 * Realign DNA; currently keeping existing gaps in introns only
794 ((Alignment) al1).alignAs(al2, false, true);
795 assertEquals("ACG---GCUCCA------ACT---", al1.getSequenceAt(0)
796 .getSequenceAsString());
797 assertEquals("---CGT---TAACGA---AGT---", al1.getSequenceAt(1)
798 .getSequenceAsString());
802 * Test aligning cdna as per protein - single sequences
804 * @throws IOException
806 @Test(groups = { "Functional" }, enabled = true)
807 // TODO review / update this test after redesign of alignAs method
808 public void testAlignAs_cdnaAsProtein_singleSequence() throws IOException
811 * simple case insert one gap
813 verifyAlignAs(">dna\nCAAaaa\n", ">protein\nQ-K\n", "CAA---aaa");
816 * simple case but with sequence offsets
818 verifyAlignAs(">dna/5-10\nCAAaaa\n", ">protein/20-21\nQ-K\n",
822 * insert gaps as per protein, drop gaps within codons
824 verifyAlignAs(">dna/10-18\nCA-Aa-aa--AGA\n", ">aa/6-8\n-Q-K--R\n",
825 "---CAA---aaa------AGA");
829 * Helper method that makes mappings and then aligns the first alignment as
835 * @throws IOException
837 public void verifyAlignAs(String fromSeqs, String toSeqs, String expected)
841 * Load alignments and add mappings from nucleotide to protein (or from
842 * first to second if both the same type)
844 AlignmentI al1 = loadAlignment(fromSeqs, FileFormat.Fasta);
845 AlignmentI al2 = loadAlignment(toSeqs, FileFormat.Fasta);
846 makeMappings(al1, al2);
849 * Realign DNA; currently keeping existing gaps in introns only
851 ((Alignment) al1).alignAs(al2, false, true);
852 assertEquals(expected, al1.getSequenceAt(0).getSequenceAsString());
856 * Helper method to make mappings between sequences, and add the mappings to
857 * the 'mapped from' alignment
862 public void makeMappings(AlignmentI alFrom, AlignmentI alTo)
864 int ratio = (alFrom.isNucleotide() == alTo.isNucleotide() ? 1 : 3);
866 AlignedCodonFrame acf = new AlignedCodonFrame();
868 for (int i = 0; i < alFrom.getHeight(); i++)
870 SequenceI seqFrom = alFrom.getSequenceAt(i);
871 SequenceI seqTo = alTo.getSequenceAt(i);
872 MapList ml = new MapList(new int[] { seqFrom.getStart(),
874 new int[] { seqTo.getStart(), seqTo.getEnd() }, ratio, 1);
875 acf.addMap(seqFrom, seqTo, ml);
879 * not sure whether mappings 'belong' or protein or nucleotide
880 * alignment, so adding to both ;~)
882 alFrom.addCodonFrame(acf);
883 alTo.addCodonFrame(acf);
887 * Test aligning dna as per protein alignment, for the case where there are
888 * introns (i.e. some dna sites have no mapping from a peptide).
890 * @throws IOException
892 @Test(groups = { "Functional" }, enabled = false)
893 // TODO review / update this test after redesign of alignAs method
894 public void testAlignAs_dnaAsProtein_withIntrons() throws IOException
897 * Load alignments and add mappings for cDNA to protein
899 String dna1 = "A-Aa-gG-GCC-cT-TT";
900 String dna2 = "c--CCGgg-TT--T-AA-A";
901 AlignmentI al1 = loadAlignment(">Dna1/6-17\n" + dna1
902 + "\n>Dna2/20-31\n" + dna2 + "\n", FileFormat.Fasta);
903 AlignmentI al2 = loadAlignment(
904 ">Pep1/7-9\n-P--YK\n>Pep2/11-13\nG-T--F\n", FileFormat.Fasta);
905 AlignedCodonFrame acf = new AlignedCodonFrame();
906 // Seq1 has intron at dna positions 3,4,9 so splice is AAG GCC TTT
907 // Seq2 has intron at dna positions 1,5,6 so splice is CCG TTT AAA
908 MapList ml1 = new MapList(new int[] { 6, 7, 10, 13, 15, 17 }, new int[]
910 acf.addMap(al1.getSequenceAt(0), al2.getSequenceAt(0), ml1);
911 MapList ml2 = new MapList(new int[] { 21, 23, 26, 31 }, new int[] { 11,
913 acf.addMap(al1.getSequenceAt(1), al2.getSequenceAt(1), ml2);
914 al2.addCodonFrame(acf);
917 * Align ignoring gaps in dna introns and exons
919 ((Alignment) al1).alignAs(al2, false, false);
920 assertEquals("---AAagG------GCCcTTT", al1.getSequenceAt(0)
921 .getSequenceAsString());
922 // note 1 gap in protein corresponds to 'gg-' in DNA (3 positions)
923 assertEquals("cCCGgg-TTT------AAA", al1.getSequenceAt(1)
924 .getSequenceAsString());
927 * Reset and realign, preserving gaps in dna introns and exons
929 al1.getSequenceAt(0).setSequence(dna1);
930 al1.getSequenceAt(1).setSequence(dna2);
931 ((Alignment) al1).alignAs(al2, true, true);
932 // String dna1 = "A-Aa-gG-GCC-cT-TT";
933 // String dna2 = "c--CCGgg-TT--T-AA-A";
934 // assumption: we include 'the greater of' protein/dna gap lengths, not both
935 assertEquals("---A-Aa-gG------GCC-cT-TT", al1.getSequenceAt(0)
936 .getSequenceAsString());
937 assertEquals("c--CCGgg-TT--T------AA-A", al1.getSequenceAt(1)
938 .getSequenceAsString());
941 @Test(groups = "Functional")
942 public void testCopyConstructor() throws IOException
944 AlignmentI protein = loadAlignment(AA_SEQS_1, FileFormat.Fasta);
945 // create sequence and alignment datasets
946 protein.setDataset(null);
947 AlignedCodonFrame acf = new AlignedCodonFrame();
948 List<AlignedCodonFrame> acfList = Arrays.asList(new AlignedCodonFrame[]
950 protein.getDataset().setCodonFrames(acfList);
951 AlignmentI copy = new Alignment(protein);
954 * copy has different aligned sequences but the same dataset sequences
956 assertFalse(copy.getSequenceAt(0) == protein.getSequenceAt(0));
957 assertFalse(copy.getSequenceAt(1) == protein.getSequenceAt(1));
958 assertSame(copy.getSequenceAt(0).getDatasetSequence(), protein
959 .getSequenceAt(0).getDatasetSequence());
960 assertSame(copy.getSequenceAt(1).getDatasetSequence(), protein
961 .getSequenceAt(1).getDatasetSequence());
963 // TODO should the copy constructor copy the dataset?
964 // or make a new one referring to the same dataset sequences??
965 assertNull(copy.getDataset());
966 // TODO test metadata is copied when AlignmentI is a dataset
968 // assertArrayEquals(copy.getDataset().getSequencesArray(), protein
969 // .getDataset().getSequencesArray());
973 * Test behaviour of createDataset
975 * @throws IOException
977 @Test(groups = "Functional")
978 public void testCreateDatasetAlignment() throws IOException
980 AlignmentI protein = new FormatAdapter().readFile(AA_SEQS_1,
981 DataSourceType.PASTE, FileFormat.Fasta);
983 * create a dataset sequence on first sequence
984 * leave the second without one
986 protein.getSequenceAt(0).createDatasetSequence();
987 assertNotNull(protein.getSequenceAt(0).getDatasetSequence());
988 assertNull(protein.getSequenceAt(1).getDatasetSequence());
991 * add a mapping to the alignment
993 AlignedCodonFrame acf = new AlignedCodonFrame();
994 protein.addCodonFrame(acf);
995 assertNull(protein.getDataset());
996 assertTrue(protein.getCodonFrames().contains(acf));
999 * create the alignment dataset
1000 * note this creates sequence datasets where missing
1001 * as a side-effect (in this case, on seq2
1003 // TODO promote this method to AlignmentI
1004 ((Alignment) protein).createDatasetAlignment();
1006 AlignmentI ds = protein.getDataset();
1008 // side-effect: dataset created on second sequence
1009 assertNotNull(protein.getSequenceAt(1).getDatasetSequence());
1010 // dataset alignment has references to dataset sequences
1011 assertEquals(ds.getSequenceAt(0), protein.getSequenceAt(0)
1012 .getDatasetSequence());
1013 assertEquals(ds.getSequenceAt(1), protein.getSequenceAt(1)
1014 .getDatasetSequence());
1016 // codon frames should have been moved to the dataset
1017 // getCodonFrames() should delegate to the dataset:
1018 assertTrue(protein.getCodonFrames().contains(acf));
1019 // prove the codon frames are indeed on the dataset:
1020 assertTrue(ds.getCodonFrames().contains(acf));
1024 * tests the addition of *all* sequences referred to by a sequence being added
1027 @Test(groups = "Functional")
1028 public void testCreateDatasetAlignmentWithMappedToSeqs()
1030 // Alignment with two sequences, gapped.
1031 SequenceI sq1 = new Sequence("sq1", "A--SDF");
1032 SequenceI sq2 = new Sequence("sq2", "G--TRQ");
1034 // cross-references to two more sequences.
1035 DBRefEntry dbr = new DBRefEntry("SQ1", "", "sq3");
1036 SequenceI sq3 = new Sequence("sq3", "VWANG");
1037 dbr.setMap(new Mapping(sq3, new MapList(new int[] { 1, 4 }, new int[] {
1041 SequenceI sq4 = new Sequence("sq4", "ERKWI");
1042 DBRefEntry dbr2 = new DBRefEntry("SQ2", "", "sq4");
1043 dbr2.setMap(new Mapping(sq4, new MapList(new int[] { 1, 4 }, new int[] {
1046 // and a 1:1 codonframe mapping between them.
1047 AlignedCodonFrame alc = new AlignedCodonFrame();
1048 alc.addMap(sq1, sq2, new MapList(new int[] { 1, 4 },
1049 new int[] { 1, 4 }, 1, 1));
1051 AlignmentI protein = new Alignment(new SequenceI[] { sq1, sq2 });
1054 * create the alignment dataset
1055 * note this creates sequence datasets where missing
1056 * as a side-effect (in this case, on seq2
1059 // TODO promote this method to AlignmentI
1060 ((Alignment) protein).createDatasetAlignment();
1062 AlignmentI ds = protein.getDataset();
1064 // should be 4 sequences in dataset - two materialised, and two propagated
1066 assertEquals(4, ds.getHeight());
1067 assertTrue(ds.getSequences().contains(sq1.getDatasetSequence()));
1068 assertTrue(ds.getSequences().contains(sq2.getDatasetSequence()));
1069 assertTrue(ds.getSequences().contains(sq3));
1070 assertTrue(ds.getSequences().contains(sq4));
1071 // Should have one codon frame mapping between sq1 and sq2 via dataset
1073 assertEquals(ds.getCodonFrame(sq1.getDatasetSequence()),
1074 ds.getCodonFrame(sq2.getDatasetSequence()));
1077 @Test(groups = "Functional")
1078 public void testAddCodonFrame()
1080 AlignmentI align = new Alignment(new SequenceI[] {});
1081 AlignedCodonFrame acf = new AlignedCodonFrame();
1082 align.addCodonFrame(acf);
1083 assertEquals(1, align.getCodonFrames().size());
1084 assertTrue(align.getCodonFrames().contains(acf));
1085 // can't add the same object twice:
1086 align.addCodonFrame(acf);
1087 assertEquals(1, align.getCodonFrames().size());
1089 // create dataset alignment - mappings move to dataset
1090 ((Alignment) align).createDatasetAlignment();
1091 assertSame(align.getCodonFrames(), align.getDataset().getCodonFrames());
1092 assertEquals(1, align.getCodonFrames().size());
1094 AlignedCodonFrame acf2 = new AlignedCodonFrame();
1095 align.addCodonFrame(acf2);
1096 assertTrue(align.getDataset().getCodonFrames().contains(acf));
1099 @Test(groups = "Functional")
1100 public void testAddSequencePreserveDatasetIntegrity()
1102 Sequence seq = new Sequence("testSeq", "ABCDEFGHIJKLMNOPQRSTUVWXYZ");
1103 Alignment align = new Alignment(new SequenceI[] { seq });
1104 align.createDatasetAlignment();
1105 AlignmentI ds = align.getDataset();
1106 SequenceI copy = new Sequence(seq);
1107 copy.insertCharAt(3, 5, '-');
1108 align.addSequence(copy);
1109 Assert.assertEquals(align.getDataset().getHeight(), 1,
1110 "Dataset shouldn't have more than one sequence.");
1112 Sequence seq2 = new Sequence("newtestSeq", "ABCDEFGHIJKLMNOPQRSTUVWXYZ");
1113 align.addSequence(seq2);
1114 Assert.assertEquals(align.getDataset().getHeight(), 2,
1115 "Dataset should now have two sequences.");
1117 assertAlignmentDatasetRefs(align,
1118 "addSequence broke dataset reference integrity");
1122 * Tests that dbrefs with mappings to sequence get updated if the sequence
1123 * acquires a dataset sequence
1125 @Test(groups = "Functional")
1126 public void testCreateDataset_updateDbrefMappings()
1128 SequenceI pep = new Sequence("pep", "ASD");
1129 SequenceI dna = new Sequence("dna", "aaaGCCTCGGATggg");
1130 SequenceI cds = new Sequence("cds", "GCCTCGGAT");
1132 // add dbref from dna to peptide
1133 DBRefEntry dbr = new DBRefEntry("UNIPROT", "", "pep");
1134 dbr.setMap(new Mapping(pep, new MapList(new int[] { 4, 15 }, new int[] {
1138 // add dbref from dna to peptide
1139 DBRefEntry dbr2 = new DBRefEntry("UNIPROT", "", "pep");
1140 dbr2.setMap(new Mapping(pep, new MapList(new int[] { 1, 12 }, new int[]
1144 // add dbref from peptide to dna
1145 DBRefEntry dbr3 = new DBRefEntry("EMBL", "", "dna");
1146 dbr3.setMap(new Mapping(dna, new MapList(new int[] { 1, 4 }, new int[] {
1150 // add dbref from peptide to cds
1151 DBRefEntry dbr4 = new DBRefEntry("EMBLCDS", "", "cds");
1152 dbr4.setMap(new Mapping(cds, new MapList(new int[] { 1, 4 }, new int[] {
1156 AlignmentI protein = new Alignment(new SequenceI[] { pep });
1159 * create the alignment dataset
1161 ((Alignment) protein).createDatasetAlignment();
1163 AlignmentI ds = protein.getDataset();
1165 // should be 3 sequences in dataset
1166 assertEquals(3, ds.getHeight());
1167 assertTrue(ds.getSequences().contains(pep.getDatasetSequence()));
1168 assertTrue(ds.getSequences().contains(dna));
1169 assertTrue(ds.getSequences().contains(cds));
1172 * verify peptide.cdsdbref.peptidedbref is now mapped to peptide dataset
1174 DBRefEntry[] dbRefs = pep.getDBRefs();
1175 assertEquals(2, dbRefs.length);
1176 assertSame(dna, dbRefs[0].map.to);
1177 assertSame(cds, dbRefs[1].map.to);
1178 assertEquals(1, dna.getDBRefs().length);
1179 assertSame(pep.getDatasetSequence(), dna.getDBRefs()[0].map.to);
1180 assertEquals(1, cds.getDBRefs().length);
1181 assertSame(pep.getDatasetSequence(), cds.getDBRefs()[0].map.to);
1184 @Test(groups = { "Functional" })
1185 public void testFindGroup()
1187 SequenceI seq1 = new Sequence("seq1", "ABCDEF---GHI");
1188 SequenceI seq2 = new Sequence("seq2", "---JKLMNO---");
1189 AlignmentI a = new Alignment(new SequenceI[] { seq1, seq2 });
1191 assertNull(a.findGroup(null, 0));
1192 assertNull(a.findGroup(seq1, 1));
1193 assertNull(a.findGroup(seq1, -1));
1196 * add a group consisting of just "DEF"
1198 SequenceGroup sg1 = new SequenceGroup();
1199 sg1.addSequence(seq1, false);
1204 assertNull(a.findGroup(seq1, 2)); // position not in group
1205 assertNull(a.findGroup(seq1, 6)); // position not in group
1206 assertNull(a.findGroup(seq2, 5)); // sequence not in group
1207 assertSame(a.findGroup(seq1, 3), sg1); // yes
1208 assertSame(a.findGroup(seq1, 4), sg1);
1209 assertSame(a.findGroup(seq1, 5), sg1);
1212 * add a group consisting of
1216 SequenceGroup sg2 = new SequenceGroup();
1217 sg2.addSequence(seq1, false);
1218 sg2.addSequence(seq2, false);
1223 assertNull(a.findGroup(seq1, 2)); // unchanged
1224 assertSame(a.findGroup(seq1, 3), sg1); // unchanged
1226 * if a residue is in more than one group, method returns
1227 * the first found (in order groups were added)
1229 assertSame(a.findGroup(seq1, 4), sg1);
1230 assertSame(a.findGroup(seq1, 5), sg1);
1233 * seq2 only belongs to the second group
1235 assertSame(a.findGroup(seq2, 4), sg2);
1236 assertSame(a.findGroup(seq2, 5), sg2);
1237 assertSame(a.findGroup(seq2, 6), sg2);
1238 assertSame(a.findGroup(seq2, 7), sg2);
1239 assertNull(a.findGroup(seq2, 3));
1240 assertNull(a.findGroup(seq2, 8));
1243 @Test(groups = { "Functional" })
1244 public void testDeleteSequenceByIndex()
1246 // create random alignment
1247 AlignmentGenerator gen = new AlignmentGenerator(false);
1248 AlignmentI a = gen.generate(20, 15, 123, 5, 5);
1250 // delete sequence 10, alignment reduced by 1
1251 int height = a.getAbsoluteHeight();
1252 a.deleteSequence(10);
1253 assertEquals(a.getAbsoluteHeight(), height - 1);
1255 // try to delete -ve index, nothing happens
1256 a.deleteSequence(-1);
1257 assertEquals(a.getAbsoluteHeight(), height - 1);
1259 // try to delete beyond end of alignment, nothing happens
1260 a.deleteSequence(14);
1261 assertEquals(a.getAbsoluteHeight(), height - 1);
1264 @Test(groups = { "Functional" })
1265 public void testDeleteSequenceBySeq()
1267 // create random alignment
1268 AlignmentGenerator gen = new AlignmentGenerator(false);
1269 AlignmentI a = gen.generate(20, 15, 123, 5, 5);
1271 // delete sequence 10, alignment reduced by 1
1272 int height = a.getAbsoluteHeight();
1273 SequenceI seq = a.getSequenceAt(10);
1274 a.deleteSequence(seq);
1275 assertEquals(a.getAbsoluteHeight(), height - 1);
1277 // try to delete non-existent sequence, nothing happens
1278 seq = new Sequence("cds", "GCCTCGGAT");
1279 assertEquals(a.getAbsoluteHeight(), height - 1);
1282 @Test(groups = { "Functional" })
1283 public void testDeleteHiddenSequence()
1285 // create random alignment
1286 AlignmentGenerator gen = new AlignmentGenerator(false);
1287 AlignmentI a = gen.generate(20, 15, 123, 5, 5);
1289 // delete a sequence which is hidden, check it is NOT removed from hidden
1291 int height = a.getAbsoluteHeight();
1292 SequenceI seq = a.getSequenceAt(2);
1293 a.getHiddenSequences().hideSequence(seq);
1294 assertEquals(a.getHiddenSequences().getSize(), 1);
1295 a.deleteSequence(2);
1296 assertEquals(a.getAbsoluteHeight(), height - 1);
1297 assertEquals(a.getHiddenSequences().getSize(), 1);
1299 // delete a sequence which is not hidden, check hiddenSequences are not
1301 a.deleteSequence(10);
1302 assertEquals(a.getAbsoluteHeight(), height - 2);
1303 assertEquals(a.getHiddenSequences().getSize(), 1);
1307 groups = "Functional",
1308 expectedExceptions = { IllegalArgumentException.class })
1309 public void testSetDataset_selfReference()
1311 SequenceI seq = new Sequence("a", "a");
1312 AlignmentI alignment = new Alignment(new SequenceI[] { seq });
1313 alignment.setDataset(alignment);
1316 @Test(groups = "Functional")
1317 public void testAppend()
1319 SequenceI seq = new Sequence("seq1", "FRMLPSRT-A--L-");
1320 AlignmentI alignment = new Alignment(new SequenceI[] { seq });
1321 alignment.setGapCharacter('-');
1322 SequenceI seq2 = new Sequence("seq1", "KP..L.FQII.");
1323 AlignmentI alignment2 = new Alignment(new SequenceI[] { seq2 });
1324 alignment2.setGapCharacter('.');
1326 alignment.append(alignment2);
1328 assertEquals('-', alignment.getGapCharacter());
1329 assertSame(seq, alignment.getSequenceAt(0));
1330 assertEquals("KP--L-FQII-", alignment.getSequenceAt(1)
1331 .getSequenceAsString());
1333 // todo test coverage for annotations, mappings, groups,
1334 // hidden sequences, properties
1338 * test that calcId == null on findOrCreate doesn't raise an NPE, and yields
1339 * an annotation with a null calcId
1342 @Test(groups = "Functional")
1343 public void testFindOrCreateForNullCalcId()
1345 SequenceI seq = new Sequence("seq1", "FRMLPSRT-A--L-");
1346 AlignmentI alignment = new Alignment(new SequenceI[] { seq });
1348 AlignmentAnnotation ala = alignment.findOrCreateAnnotation(
1349 "Temperature Factor", null, false, seq, null);
1351 assertEquals(seq, ala.sequenceRef);
1352 assertEquals("", ala.getCalcId());
1355 @Test(groups = "Functional")
1356 public void testPropagateInsertions()
1358 // create an alignment with no gaps - this will be the profile seq and other
1360 AlignmentGenerator gen = new AlignmentGenerator(false);
1361 AlignmentI al = gen.generate(25, 10, 1234, 0, 0);
1363 // get the profileseq
1364 SequenceI profileseq = al.getSequenceAt(0);
1365 SequenceI gappedseq = new Sequence(profileseq);
1366 gappedseq.insertCharAt(5, al.getGapCharacter());
1367 gappedseq.insertCharAt(6, al.getGapCharacter());
1368 gappedseq.insertCharAt(7, al.getGapCharacter());
1369 gappedseq.insertCharAt(8, al.getGapCharacter());
1371 // force different kinds of padding
1372 al.getSequenceAt(3).deleteChars(2, 23);
1373 al.getSequenceAt(4).deleteChars(2, 27);
1374 al.getSequenceAt(5).deleteChars(10, 27);
1376 // create an alignment view with the gapped sequence
1377 SequenceI[] seqs = new SequenceI[1];
1378 seqs[0] = gappedseq;
1379 AlignmentI newal = new Alignment(seqs);
1380 HiddenColumns hidden = new HiddenColumns();
1381 hidden.hideColumns(15, 17);
1383 AlignmentView view = new AlignmentView(newal, hidden, null, true, false,
1386 // confirm that original contigs are as expected
1387 Iterator<int[]> visible = hidden.getVisContigsIterator(0, 25, false);
1388 int[] region = visible.next();
1389 assertEquals("[0, 14]", Arrays.toString(region));
1390 region = visible.next();
1391 assertEquals("[18, 24]", Arrays.toString(region));
1393 // propagate insertions
1394 HiddenColumns result = al.propagateInsertions(profileseq, view);
1396 // confirm that the contigs have changed to account for the gaps
1397 visible = result.getVisContigsIterator(0, 25, false);
1398 region = visible.next();
1399 assertEquals("[0, 10]", Arrays.toString(region));
1400 region = visible.next();
1401 assertEquals("[14, 24]", Arrays.toString(region));
1403 // confirm the alignment has been changed so that the other sequences have
1404 // gaps inserted where the columns are hidden
1405 assertFalse(Comparison.isGap(al.getSequenceAt(1).getSequence()[10]));
1406 assertTrue(Comparison.isGap(al.getSequenceAt(1).getSequence()[11]));
1407 assertTrue(Comparison.isGap(al.getSequenceAt(1).getSequence()[12]));
1408 assertTrue(Comparison.isGap(al.getSequenceAt(1).getSequence()[13]));
1409 assertFalse(Comparison.isGap(al.getSequenceAt(1).getSequence()[14]));
1413 @Test(groups = "Functional")
1414 public void testPropagateInsertionsOverlap()
1416 // test propagateInsertions where gaps and hiddenColumns overlap
1418 // create an alignment with no gaps - this will be the profile seq and other
1420 AlignmentGenerator gen = new AlignmentGenerator(false);
1421 AlignmentI al = gen.generate(20, 10, 1234, 0, 0);
1423 // get the profileseq
1424 SequenceI profileseq = al.getSequenceAt(0);
1425 SequenceI gappedseq = new Sequence(profileseq);
1426 gappedseq.insertCharAt(5, al.getGapCharacter());
1427 gappedseq.insertCharAt(6, al.getGapCharacter());
1428 gappedseq.insertCharAt(7, al.getGapCharacter());
1429 gappedseq.insertCharAt(8, al.getGapCharacter());
1431 // create an alignment view with the gapped sequence
1432 SequenceI[] seqs = new SequenceI[1];
1433 seqs[0] = gappedseq;
1434 AlignmentI newal = new Alignment(seqs);
1436 // hide columns so that some overlap with the gaps
1437 HiddenColumns hidden = new HiddenColumns();
1438 hidden.hideColumns(7, 10);
1440 AlignmentView view = new AlignmentView(newal, hidden, null, true, false,
1443 // confirm that original contigs are as expected
1444 Iterator<int[]> visible = hidden.getVisContigsIterator(0, 20, false);
1445 int[] region = visible.next();
1446 assertEquals("[0, 6]", Arrays.toString(region));
1447 region = visible.next();
1448 assertEquals("[11, 19]", Arrays.toString(region));
1449 assertFalse(visible.hasNext());
1451 // propagate insertions
1452 HiddenColumns result = al.propagateInsertions(profileseq, view);
1454 // confirm that the contigs have changed to account for the gaps
1455 visible = result.getVisContigsIterator(0, 20, false);
1456 region = visible.next();
1457 assertEquals("[0, 4]", Arrays.toString(region));
1458 region = visible.next();
1459 assertEquals("[7, 19]", Arrays.toString(region));
1460 assertFalse(visible.hasNext());
1462 // confirm the alignment has been changed so that the other sequences have
1463 // gaps inserted where the columns are hidden
1464 assertFalse(Comparison.isGap(al.getSequenceAt(1).getSequence()[4]));
1465 assertTrue(Comparison.isGap(al.getSequenceAt(1).getSequence()[5]));
1466 assertTrue(Comparison.isGap(al.getSequenceAt(1).getSequence()[6]));
1467 assertFalse(Comparison.isGap(al.getSequenceAt(1).getSequence()[7]));
1470 @Test(groups = { "Functional" })
1471 public void testPadGaps()
1473 SequenceI seq1 = new Sequence("seq1", "ABCDEF--");
1474 SequenceI seq2 = new Sequence("seq2", "-JKLMNO--");
1475 SequenceI seq3 = new Sequence("seq2", "-PQR");
1476 AlignmentI a = new Alignment(new SequenceI[] { seq1, seq2, seq3 });
1477 a.setGapCharacter('.'); // this replaces existing gaps
1478 assertEquals("ABCDEF..", seq1.getSequenceAsString());
1480 // trailing gaps are pruned, short sequences padded with gap character
1481 assertEquals("ABCDEF.", seq1.getSequenceAsString());
1482 assertEquals(".JKLMNO", seq2.getSequenceAsString());
1483 assertEquals(".PQR...", seq3.getSequenceAsString());