X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=test%2Fjalview%2Fanalysis%2FAlignmentUtilsTests.java;h=34ec73b15b8c72558c057ac89e141076852eb0f7;hb=fb91c5557602ff6581e1b3c1333f1c0ee7a02a5e;hp=f9c1a11335a7a56bb9d8b5d089b6f1ace33e9f57;hpb=4ffd862cba27561e83a5e25982c742627cc363e5;p=jalview.git diff --git a/test/jalview/analysis/AlignmentUtilsTests.java b/test/jalview/analysis/AlignmentUtilsTests.java index f9c1a11..34ec73b 100644 --- a/test/jalview/analysis/AlignmentUtilsTests.java +++ b/test/jalview/analysis/AlignmentUtilsTests.java @@ -22,10 +22,12 @@ package jalview.analysis; import static org.testng.AssertJUnit.assertEquals; import static org.testng.AssertJUnit.assertFalse; +import static org.testng.AssertJUnit.assertNotNull; import static org.testng.AssertJUnit.assertNull; import static org.testng.AssertJUnit.assertSame; import static org.testng.AssertJUnit.assertTrue; +import jalview.analysis.AlignmentUtils.DnaVariant; import jalview.datamodel.AlignedCodonFrame; import jalview.datamodel.Alignment; import jalview.datamodel.AlignmentAnnotation; @@ -46,51 +48,15 @@ import jalview.util.MappingUtils; import java.io.IOException; import java.util.ArrayList; import java.util.Arrays; -import java.util.HashSet; -import java.util.Iterator; +import java.util.LinkedHashMap; import java.util.List; import java.util.Map; -import java.util.Set; +import java.util.TreeMap; import org.testng.annotations.Test; public class AlignmentUtilsTests { - // @formatter:off - private static final String TEST_DATA = - "# STOCKHOLM 1.0\n" + - "#=GS D.melanogaster.1 AC AY119185.1/838-902\n" + - "#=GS D.melanogaster.2 AC AC092237.1/57223-57161\n" + - "#=GS D.melanogaster.3 AC AY060611.1/560-627\n" + - "D.melanogaster.1 G.AGCC.CU...AUGAUCGA\n" + - "#=GR D.melanogaster.1 SS ................((((\n" + - "D.melanogaster.2 C.AUUCAACU.UAUGAGGAU\n" + - "#=GR D.melanogaster.2 SS ................((((\n" + - "D.melanogaster.3 G.UGGCGCU..UAUGACGCA\n" + - "#=GR D.melanogaster.3 SS (.(((...(....(((((((\n" + - "//"; - - private static final String AA_SEQS_1 = - ">Seq1Name\n" + - "K-QY--L\n" + - ">Seq2Name\n" + - "-R-FP-W-\n"; - - private static final String CDNA_SEQS_1 = - ">Seq1Name\n" + - "AC-GG--CUC-CAA-CT\n" + - ">Seq2Name\n" + - "-CG-TTA--ACG---AAGT\n"; - - private static final String CDNA_SEQS_2 = - ">Seq1Name\n" + - "GCTCGUCGTACT\n" + - ">Seq2Name\n" + - "GGGTCAGGCAGT\n"; - // @formatter:on - - // public static Sequence ts=new - // Sequence("short","ASDASDASDASDASDASDASDASDASDASDASDASDASD"); public static Sequence ts = new Sequence("short", "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklm"); @@ -497,30 +463,6 @@ public class AlignmentUtilsTests } /** - * Test for the method that generates an aligned translated sequence from one - * mapping. - */ - @Test(groups = { "Functional" }) - public void testGetAlignedTranslation_dnaLikeProtein() - { - // dna alignment will be replaced - SequenceI dna = new Sequence("Seq1", "T-G-CC-A--T-TAC-CAG-"); - dna.createDatasetSequence(); - // protein alignment will be 'applied' to dna - SequenceI protein = new Sequence("Seq1", "-CH-Y--Q-"); - protein.createDatasetSequence(); - MapList map = new MapList(new int[] { 1, 12 }, new int[] { 1, 4 }, 3, 1); - AlignedCodonFrame acf = new AlignedCodonFrame(); - acf.addMap(dna.getDatasetSequence(), protein.getDatasetSequence(), map); - - final SequenceI aligned = AlignmentUtils.getAlignedTranslation(protein, - '-', acf); - assertEquals("---TGCCAT---TAC------CAG---", - aligned.getSequenceAsString()); - assertSame(aligned.getDatasetSequence(), dna.getDatasetSequence()); - } - - /** * Test the method that realigns protein to match mapped codon alignment. */ @Test(groups = { "Functional" }) @@ -571,9 +513,15 @@ public class AlignmentUtilsTests @Test(groups = { "Functional" }) public void testTranslatesAs() { + // null arguments check + assertFalse(AlignmentUtils.translatesAs(null, 0, null)); + assertFalse(AlignmentUtils.translatesAs(new char[] { 't' }, 0, null)); + assertFalse(AlignmentUtils.translatesAs(null, 0, new char[] { 'a' })); + + // straight translation assertTrue(AlignmentUtils.translatesAs("tttcccaaaggg".toCharArray(), 0, "FPKG".toCharArray())); - // with start codon (not in protein) + // with extra start codon (not in protein) assertTrue(AlignmentUtils.translatesAs("atgtttcccaaaggg".toCharArray(), 3, "FPKG".toCharArray())); // with stop codon1 (not in protein) @@ -601,7 +549,7 @@ public class AlignmentUtilsTests assertTrue(AlignmentUtils.translatesAs( "atgtttcccaaagggtga".toCharArray(), 3, "FPKG".toCharArray())); - // with embedded stop codon + // with embedded stop codons assertTrue(AlignmentUtils.translatesAs( "atgtttTAGcccaaaTAAgggtga".toCharArray(), 3, "F*PK*G".toCharArray())); @@ -609,6 +557,26 @@ public class AlignmentUtilsTests // wrong protein assertFalse(AlignmentUtils.translatesAs("tttcccaaaggg".toCharArray(), 0, "FPMG".toCharArray())); + + // truncated dna + assertFalse(AlignmentUtils.translatesAs("tttcccaaagg".toCharArray(), 0, + "FPKG".toCharArray())); + + // truncated protein + assertFalse(AlignmentUtils.translatesAs("tttcccaaaggg".toCharArray(), + 0, "FPK".toCharArray())); + + // overlong dna (doesn't end in stop codon) + assertFalse(AlignmentUtils.translatesAs( + "tttcccaaagggttt".toCharArray(), 0, "FPKG".toCharArray())); + + // dna + stop codon + more + assertFalse(AlignmentUtils.translatesAs( + "tttcccaaagggttaga".toCharArray(), 0, "FPKG".toCharArray())); + + // overlong protein + assertFalse(AlignmentUtils.translatesAs("tttcccaaaggg".toCharArray(), + 0, "FPKGQ".toCharArray())); } /** @@ -1006,149 +974,175 @@ public class AlignmentUtilsTests @Test(groups = { "Functional" }) public void testMakeCdsAlignment() { + /* + * scenario: + * dna1 --> [4, 6] [10,12] --> pep1 + * dna2 --> [1, 3] [7, 9] [13,15] --> pep1 + */ SequenceI dna1 = new Sequence("dna1", "aaaGGGcccTTTaaa"); SequenceI dna2 = new Sequence("dna2", "GGGcccTTTaaaCCC"); SequenceI pep1 = new Sequence("pep1", "GF"); SequenceI pep2 = new Sequence("pep2", "GFP"); + pep1.addDBRef(new DBRefEntry("UNIPROT", "0", "pep1")); + pep2.addDBRef(new DBRefEntry("UNIPROT", "0", "pep2")); dna1.createDatasetSequence(); dna2.createDatasetSequence(); pep1.createDatasetSequence(); pep2.createDatasetSequence(); - dna1.addSequenceFeature(new SequenceFeature("CDS", "cds1", 4, 6, 0f, - null)); - dna1.addSequenceFeature(new SequenceFeature("CDS", "cds2", 10, 12, 0f, - null)); - dna2.addSequenceFeature(new SequenceFeature("CDS", "cds3", 1, 3, 0f, - null)); - dna2.addSequenceFeature(new SequenceFeature("CDS", "cds4", 7, 9, 0f, - null)); - dna2.addSequenceFeature(new SequenceFeature("CDS", "cds5", 13, 15, 0f, - null)); + AlignmentI dna = new Alignment(new SequenceI[] { dna1, dna2 }); + dna.setDataset(null); + + /* + * need a sourceDbRef if we are to construct dbrefs to the CDS + * sequence + */ + DBRefEntry dbref = new DBRefEntry("ENSEMBL", "0", "dna1"); + dna1.getDatasetSequence().addDBRef(dbref); + org.testng.Assert.assertEquals(dbref, dna1.getPrimaryDBRefs().get(0)); + dbref = new DBRefEntry("ENSEMBL", "0", "dna2"); + dna2.getDatasetSequence().addDBRef(dbref); + org.testng.Assert.assertEquals(dbref, dna2.getPrimaryDBRefs().get(0)); - List mappings = new ArrayList(); + /* + * CDS sequences are 'discovered' from dna-to-protein mappings on the alignment + * dataset (e.g. added from dbrefs by CrossRef.findXrefSequences) + */ MapList map = new MapList(new int[] { 4, 6, 10, 12 }, new int[] { 1, 2 }, 3, 1); AlignedCodonFrame acf = new AlignedCodonFrame(); acf.addMap(dna1.getDatasetSequence(), pep1.getDatasetSequence(), map); - mappings.add(acf); + dna.addCodonFrame(acf); map = new MapList(new int[] { 1, 3, 7, 9, 13, 15 }, new int[] { 1, 3 }, 3, 1); acf = new AlignedCodonFrame(); acf.addMap(dna2.getDatasetSequence(), pep2.getDatasetSequence(), map); - mappings.add(acf); + dna.addCodonFrame(acf); + /* + * execute method under test: + */ AlignmentI cds = AlignmentUtils.makeCdsAlignment(new SequenceI[] { - dna1, dna2 }, mappings, '-'); + dna1, dna2 }, dna.getDataset(), null); + + /* + * verify cds sequences + */ assertEquals(2, cds.getSequences().size()); - assertEquals("---GGG---TTT---", cds.getSequenceAt(0) - .getSequenceAsString()); - assertEquals("GGG---TTT---CCC", cds.getSequenceAt(1) - .getSequenceAsString()); + assertEquals("GGGTTT", cds.getSequenceAt(0).getSequenceAsString()); + assertEquals("GGGTTTCCC", cds.getSequenceAt(1).getSequenceAsString()); + + /* + * verify shared, extended alignment dataset + */ + assertSame(dna.getDataset(), cds.getDataset()); + SequenceI cds1Dss = cds.getSequenceAt(0).getDatasetSequence(); + SequenceI cds2Dss = cds.getSequenceAt(1).getDatasetSequence(); + assertTrue(dna.getDataset().getSequences().contains(cds1Dss)); + assertTrue(dna.getDataset().getSequences().contains(cds2Dss)); + + /* + * verify CDS has a dbref with mapping to peptide + */ + assertNotNull(cds1Dss.getDBRefs()); + assertEquals(1, cds1Dss.getDBRefs().length); + dbref = cds1Dss.getDBRefs()[0]; + assertEquals("UNIPROT", dbref.getSource()); + assertEquals("0", dbref.getVersion()); + assertEquals("pep1", dbref.getAccessionId()); + assertNotNull(dbref.getMap()); + assertSame(pep1.getDatasetSequence(), dbref.getMap().getTo()); + MapList cdsMapping = new MapList(new int[] { 1, 6 }, + new int[] { 1, 2 }, 3, 1); + assertEquals(cdsMapping, dbref.getMap().getMap()); + + /* + * verify peptide has added a dbref with reverse mapping to CDS + */ + assertNotNull(pep1.getDBRefs()); + // FIXME pep1.getDBRefs() is 1 - is that the correct behaviour ? + assertEquals(2, pep1.getDBRefs().length); + dbref = pep1.getDBRefs()[1]; + assertEquals("ENSEMBL", dbref.getSource()); + assertEquals("0", dbref.getVersion()); + assertEquals("CDS|dna1", dbref.getAccessionId()); + assertNotNull(dbref.getMap()); + assertSame(cds1Dss, dbref.getMap().getTo()); + assertEquals(cdsMapping.getInverse(), dbref.getMap().getMap()); /* - * Verify updated mappings + * Verify mappings from CDS to peptide, cDNA to CDS, and cDNA to peptide + * the mappings are on the shared alignment dataset + * 6 mappings, 2*(DNA->CDS), 2*(DNA->Pep), 2*(CDS->Pep) */ - assertEquals(2, mappings.size()); + List cdsMappings = cds.getDataset().getCodonFrames(); + assertEquals(6, cdsMappings.size()); /* + * verify that mapping sets for dna and cds alignments are different + * [not current behaviour - all mappings are on the alignment dataset] + */ + // select -> subselect type to test. + // Assert.assertNotSame(dna.getCodonFrames(), cds.getCodonFrames()); + // assertEquals(4, dna.getCodonFrames().size()); + // assertEquals(4, cds.getCodonFrames().size()); + + /* + * Two mappings involve pep1 (dna to pep1, cds to pep1) * Mapping from pep1 to GGGTTT in first new exon sequence */ - List pep1Mapping = MappingUtils - .findMappingsForSequence(pep1, mappings); - assertEquals(1, pep1Mapping.size()); + List pep1Mappings = MappingUtils + .findMappingsForSequence(pep1, cdsMappings); + assertEquals(2, pep1Mappings.size()); + List mappings = MappingUtils + .findMappingsForSequence(cds.getSequenceAt(0), pep1Mappings); + assertEquals(1, mappings.size()); + // map G to GGG SearchResults sr = MappingUtils.buildSearchResults(pep1, 1, mappings); assertEquals(1, sr.getResults().size()); Match m = sr.getResults().get(0); - assertSame(cds.getSequenceAt(0).getDatasetSequence(), - m.getSequence()); + assertSame(cds1Dss, m.getSequence()); assertEquals(1, m.getStart()); assertEquals(3, m.getEnd()); // map F to TTT sr = MappingUtils.buildSearchResults(pep1, 2, mappings); m = sr.getResults().get(0); - assertSame(cds.getSequenceAt(0).getDatasetSequence(), - m.getSequence()); + assertSame(cds1Dss, m.getSequence()); assertEquals(4, m.getStart()); assertEquals(6, m.getEnd()); /* - * Mapping from pep2 to GGGTTTCCC in second new exon sequence + * Two mappings involve pep2 (dna to pep2, cds to pep2) + * Verify mapping from pep2 to GGGTTTCCC in second new exon sequence */ - List pep2Mapping = MappingUtils - .findMappingsForSequence(pep2, mappings); - assertEquals(1, pep2Mapping.size()); + List pep2Mappings = MappingUtils + .findMappingsForSequence(pep2, cdsMappings); + assertEquals(2, pep2Mappings.size()); + mappings = MappingUtils.findMappingsForSequence(cds.getSequenceAt(1), + pep2Mappings); + assertEquals(1, mappings.size()); // map G to GGG sr = MappingUtils.buildSearchResults(pep2, 1, mappings); assertEquals(1, sr.getResults().size()); m = sr.getResults().get(0); - assertSame(cds.getSequenceAt(1).getDatasetSequence(), - m.getSequence()); + assertSame(cds2Dss, m.getSequence()); assertEquals(1, m.getStart()); assertEquals(3, m.getEnd()); // map F to TTT sr = MappingUtils.buildSearchResults(pep2, 2, mappings); m = sr.getResults().get(0); - assertSame(cds.getSequenceAt(1).getDatasetSequence(), - m.getSequence()); + assertSame(cds2Dss, m.getSequence()); assertEquals(4, m.getStart()); assertEquals(6, m.getEnd()); // map P to CCC sr = MappingUtils.buildSearchResults(pep2, 3, mappings); m = sr.getResults().get(0); - assertSame(cds.getSequenceAt(1).getDatasetSequence(), - m.getSequence()); + assertSame(cds2Dss, m.getSequence()); assertEquals(7, m.getStart()); assertEquals(9, m.getEnd()); } /** - * Test the method that makes a cds-only sequence from a DNA sequence and its - * product mapping. Test includes the expected case that the DNA sequence - * already has a protein product (Uniprot translation) which in turn has an - * x-ref to the EMBLCDS record. - */ - @Test(groups = { "Functional" }) - public void testMakeCdsSequences() - { - SequenceI dna1 = new Sequence("dna1", "aaaGGGcccTTTaaa"); - SequenceI pep1 = new Sequence("pep1", "GF"); - dna1.createDatasetSequence(); - pep1.createDatasetSequence(); - pep1.getDatasetSequence().addDBRef( - new DBRefEntry("EMBLCDS", "2", "A12345")); - - /* - * Make the mapping from dna to protein. The protein sequence has a DBRef to - * EMBLCDS|A12345. - */ - Set mappings = new HashSet(); - MapList map = new MapList(new int[] { 4, 6, 10, 12 }, - new int[] { 1, 2 }, 3, 1); - AlignedCodonFrame acf = new AlignedCodonFrame(); - acf.addMap(dna1.getDatasetSequence(), pep1.getDatasetSequence(), map); - mappings.add(acf); - - AlignedCodonFrame newMapping = new AlignedCodonFrame(); - List ungappedColumns = new ArrayList(); - ungappedColumns.add(new int[] { 4, 6 }); - ungappedColumns.add(new int[] { 10, 12 }); - List cdsSeqs = AlignmentUtils.makeCdsSequences(dna1, acf, - ungappedColumns, - newMapping, '-'); - assertEquals(1, cdsSeqs.size()); - SequenceI cdsSeq = cdsSeqs.get(0); - - assertEquals("GGGTTT", cdsSeq.getSequenceAsString()); - assertEquals("dna1|A12345", cdsSeq.getName()); - assertEquals(1, cdsSeq.getDBRefs().length); - DBRefEntry cdsRef = cdsSeq.getDBRefs()[0]; - assertEquals("EMBLCDS", cdsRef.getSource()); - assertEquals("2", cdsRef.getVersion()); - assertEquals("A12345", cdsRef.getAccessionId()); - } - - /** * Test the method that makes a cds-only alignment from a DNA sequence and its * product mappings, for the case where there are multiple exon mappings to * different protein products. @@ -1164,18 +1158,6 @@ public class AlignmentUtilsTests pep1.createDatasetSequence(); pep2.createDatasetSequence(); pep3.createDatasetSequence(); - dna1.addSequenceFeature(new SequenceFeature("CDS", "cds1", 4, 6, 0f, - null)); - dna1.addSequenceFeature(new SequenceFeature("CDS", "cds2", 10, 12, 0f, - null)); - dna1.addSequenceFeature(new SequenceFeature("CDS", "cds3", 1, 3, 0f, - null)); - dna1.addSequenceFeature(new SequenceFeature("CDS", "cds4", 7, 9, 0f, - null)); - dna1.addSequenceFeature(new SequenceFeature("CDS", "cds5", 1, 3, 0f, - null)); - dna1.addSequenceFeature(new SequenceFeature("CDS", "cds6", 10, 12, 0f, - null)); pep1.getDatasetSequence().addDBRef( new DBRefEntry("EMBLCDS", "2", "A12345")); pep2.getDatasetSequence().addDBRef( @@ -1184,110 +1166,161 @@ public class AlignmentUtilsTests new DBRefEntry("EMBLCDS", "4", "A12347")); /* + * Create the CDS alignment + */ + AlignmentI dna = new Alignment(new SequenceI[] { dna1 }); + dna.setDataset(null); + + /* * Make the mappings from dna to protein */ - List mappings = new ArrayList(); // map ...GGG...TTT to GF MapList map = new MapList(new int[] { 4, 6, 10, 12 }, new int[] { 1, 2 }, 3, 1); AlignedCodonFrame acf = new AlignedCodonFrame(); acf.addMap(dna1.getDatasetSequence(), pep1.getDatasetSequence(), map); - mappings.add(acf); + dna.addCodonFrame(acf); // map aaa...ccc to KP map = new MapList(new int[] { 1, 3, 7, 9 }, new int[] { 1, 2 }, 3, 1); acf = new AlignedCodonFrame(); acf.addMap(dna1.getDatasetSequence(), pep2.getDatasetSequence(), map); - mappings.add(acf); + dna.addCodonFrame(acf); // map aaa......TTT to KF map = new MapList(new int[] { 1, 3, 10, 12 }, new int[] { 1, 2 }, 3, 1); acf = new AlignedCodonFrame(); acf.addMap(dna1.getDatasetSequence(), pep3.getDatasetSequence(), map); - mappings.add(acf); + dna.addCodonFrame(acf); /* - * Create the Exon alignment; also replaces the dna-to-protein mappings with - * exon-to-protein and exon-to-dna mappings + * execute method under test */ - AlignmentI exal = AlignmentUtils.makeCdsAlignment( - new SequenceI[] { dna1 }, mappings, '-'); + AlignmentI cdsal = AlignmentUtils.makeCdsAlignment( + new SequenceI[] { dna1 }, dna.getDataset(), null); /* * Verify we have 3 cds sequences, mapped to pep1/2/3 respectively */ - List cds = exal.getSequences(); + List cds = cdsal.getSequences(); assertEquals(3, cds.size()); + /* + * verify shared, extended alignment dataset + */ + assertSame(cdsal.getDataset(), dna.getDataset()); + assertTrue(dna.getDataset().getSequences() + .contains(cds.get(0).getDatasetSequence())); + assertTrue(dna.getDataset().getSequences() + .contains(cds.get(1).getDatasetSequence())); + assertTrue(dna.getDataset().getSequences() + .contains(cds.get(2).getDatasetSequence())); + + /* + * verify aligned cds sequences and their xrefs + */ SequenceI cdsSeq = cds.get(0); - assertEquals("---GGG---TTT", cdsSeq.getSequenceAsString()); - assertEquals("dna1|A12345", cdsSeq.getName()); - assertEquals(1, cdsSeq.getDBRefs().length); - DBRefEntry cdsRef = cdsSeq.getDBRefs()[0]; - assertEquals("EMBLCDS", cdsRef.getSource()); - assertEquals("2", cdsRef.getVersion()); - assertEquals("A12345", cdsRef.getAccessionId()); + assertEquals("GGGTTT", cdsSeq.getSequenceAsString()); + // assertEquals("dna1|A12345", cdsSeq.getName()); + assertEquals("CDS|dna1", cdsSeq.getName()); + // assertEquals(1, cdsSeq.getDBRefs().length); + // DBRefEntry cdsRef = cdsSeq.getDBRefs()[0]; + // assertEquals("EMBLCDS", cdsRef.getSource()); + // assertEquals("2", cdsRef.getVersion()); + // assertEquals("A12345", cdsRef.getAccessionId()); cdsSeq = cds.get(1); - assertEquals("aaa---ccc---", cdsSeq.getSequenceAsString()); - assertEquals("dna1|A12346", cdsSeq.getName()); - assertEquals(1, cdsSeq.getDBRefs().length); - cdsRef = cdsSeq.getDBRefs()[0]; - assertEquals("EMBLCDS", cdsRef.getSource()); - assertEquals("3", cdsRef.getVersion()); - assertEquals("A12346", cdsRef.getAccessionId()); + assertEquals("aaaccc", cdsSeq.getSequenceAsString()); + // assertEquals("dna1|A12346", cdsSeq.getName()); + assertEquals("CDS|dna1", cdsSeq.getName()); + // assertEquals(1, cdsSeq.getDBRefs().length); + // cdsRef = cdsSeq.getDBRefs()[0]; + // assertEquals("EMBLCDS", cdsRef.getSource()); + // assertEquals("3", cdsRef.getVersion()); + // assertEquals("A12346", cdsRef.getAccessionId()); cdsSeq = cds.get(2); - assertEquals("aaa------TTT", cdsSeq.getSequenceAsString()); - assertEquals("dna1|A12347", cdsSeq.getName()); - assertEquals(1, cdsSeq.getDBRefs().length); - cdsRef = cdsSeq.getDBRefs()[0]; - assertEquals("EMBLCDS", cdsRef.getSource()); - assertEquals("4", cdsRef.getVersion()); - assertEquals("A12347", cdsRef.getAccessionId()); + assertEquals("aaaTTT", cdsSeq.getSequenceAsString()); + // assertEquals("dna1|A12347", cdsSeq.getName()); + assertEquals("CDS|dna1", cdsSeq.getName()); + // assertEquals(1, cdsSeq.getDBRefs().length); + // cdsRef = cdsSeq.getDBRefs()[0]; + // assertEquals("EMBLCDS", cdsRef.getSource()); + // assertEquals("4", cdsRef.getVersion()); + // assertEquals("A12347", cdsRef.getAccessionId()); /* * Verify there are mappings from each cds sequence to its protein product * and also to its dna source */ - Iterator newMappingsIterator = mappings.iterator(); + List newMappings = cdsal.getCodonFrames(); - // mappings for dna1 - exon1 - pep1 - AlignedCodonFrame cdsMapping = newMappingsIterator.next(); - List dnaMappings = cdsMapping.getMappingsForSequence(dna1); - assertEquals(1, dnaMappings.size()); - assertSame(cds.get(0).getDatasetSequence(), dnaMappings.get(0) - .getTo()); - assertEquals("G(1) in CDS should map to G(4) in DNA", 4, dnaMappings - .get(0).getMap().getToPosition(1)); - List peptideMappings = cdsMapping - .getMappingsForSequence(pep1); - assertEquals(1, peptideMappings.size()); - assertSame(pep1.getDatasetSequence(), peptideMappings.get(0).getTo()); - - // mappings for dna1 - cds2 - pep2 - cdsMapping = newMappingsIterator.next(); - dnaMappings = cdsMapping.getMappingsForSequence(dna1); - assertEquals(1, dnaMappings.size()); - assertSame(cds.get(1).getDatasetSequence(), dnaMappings.get(0) - .getTo()); - assertEquals("c(4) in CDS should map to c(7) in DNA", 7, dnaMappings - .get(0).getMap().getToPosition(4)); - peptideMappings = cdsMapping.getMappingsForSequence(pep2); - assertEquals(1, peptideMappings.size()); - assertSame(pep2.getDatasetSequence(), peptideMappings.get(0).getTo()); - - // mappings for dna1 - cds3 - pep3 - cdsMapping = newMappingsIterator.next(); - dnaMappings = cdsMapping.getMappingsForSequence(dna1); - assertEquals(1, dnaMappings.size()); - assertSame(cds.get(2).getDatasetSequence(), dnaMappings.get(0) + /* + * 6 mappings involve dna1 (to pep1/2/3, cds1/2/3) + */ + List dnaMappings = MappingUtils + .findMappingsForSequence(dna1, newMappings); + assertEquals(6, dnaMappings.size()); + + /* + * dna1 to pep1 + */ + List mappings = MappingUtils + .findMappingsForSequence(pep1, dnaMappings); + assertEquals(1, mappings.size()); + assertEquals(1, mappings.get(0).getMappings().size()); + assertSame(pep1.getDatasetSequence(), mappings.get(0).getMappings() + .get(0).getMapping().getTo()); + + /* + * dna1 to cds1 + */ + List dnaToCds1Mappings = MappingUtils + .findMappingsForSequence(cds.get(0), dnaMappings); + Mapping mapping = dnaToCds1Mappings.get(0).getMappings().get(0) + .getMapping(); + assertSame(cds.get(0).getDatasetSequence(), mapping .getTo()); - assertEquals("T(4) in CDS should map to T(10) in DNA", 10, dnaMappings - .get(0).getMap().getToPosition(4)); - peptideMappings = cdsMapping.getMappingsForSequence(pep3); - assertEquals(1, peptideMappings.size()); - assertSame(pep3.getDatasetSequence(), peptideMappings.get(0).getTo()); + assertEquals("G(1) in CDS should map to G(4) in DNA", 4, mapping + .getMap().getToPosition(1)); + + /* + * dna1 to pep2 + */ + mappings = MappingUtils.findMappingsForSequence(pep2, dnaMappings); + assertEquals(1, mappings.size()); + assertEquals(1, mappings.get(0).getMappings().size()); + assertSame(pep2.getDatasetSequence(), mappings.get(0).getMappings() + .get(0).getMapping().getTo()); + + /* + * dna1 to cds2 + */ + List dnaToCds2Mappings = MappingUtils + .findMappingsForSequence(cds.get(1), dnaMappings); + mapping = dnaToCds2Mappings.get(0).getMappings().get(0).getMapping(); + assertSame(cds.get(1).getDatasetSequence(), mapping.getTo()); + assertEquals("c(4) in CDS should map to c(7) in DNA", 7, mapping + .getMap().getToPosition(4)); + + /* + * dna1 to pep3 + */ + mappings = MappingUtils.findMappingsForSequence(pep3, dnaMappings); + assertEquals(1, mappings.size()); + assertEquals(1, mappings.get(0).getMappings().size()); + assertSame(pep3.getDatasetSequence(), mappings.get(0).getMappings() + .get(0).getMapping().getTo()); + + /* + * dna1 to cds3 + */ + List dnaToCds3Mappings = MappingUtils + .findMappingsForSequence(cds.get(2), dnaMappings); + mapping = dnaToCds3Mappings.get(0).getMappings().get(0).getMapping(); + assertSame(cds.get(2).getDatasetSequence(), mapping.getTo()); + assertEquals("T(4) in CDS should map to T(10) in DNA", 10, mapping + .getMap().getToPosition(4)); } @Test(groups = { "Functional" }) @@ -1315,7 +1348,7 @@ public class AlignmentUtilsTests * @throws IOException */ @Test(groups = { "Functional" }) - public void testMapProteinSequenceToCdna_forSubsequence() + public void testMapCdnaToProtein_forSubsequence() throws IOException { SequenceI prot = new Sequence("UNIPROT|V12345", "E-I--Q", 10, 12); @@ -1324,7 +1357,7 @@ public class AlignmentUtilsTests SequenceI dna = new Sequence("EMBL|A33333", "GAA--AT-C-CAG", 40, 48); dna.createDatasetSequence(); - MapList map = AlignmentUtils.mapProteinSequenceToCdna(prot, dna); + MapList map = AlignmentUtils.mapCdnaToProtein(prot, dna); assertEquals(10, map.getToLowest()); assertEquals(12, map.getToHighest()); assertEquals(40, map.getFromLowest()); @@ -1536,168 +1569,108 @@ public class AlignmentUtilsTests dna3.createDatasetSequence(); pep1.createDatasetSequence(); pep2.createDatasetSequence(); - dna1.addSequenceFeature(new SequenceFeature("CDS", "cds1", 4, 8, 0f, - null)); - dna1.addSequenceFeature(new SequenceFeature("CDS", "cds2", 9, 12, 0f, - null)); - dna1.addSequenceFeature(new SequenceFeature("CDS", "cds3", 16, 18, 0f, - null)); - dna2.addSequenceFeature(new SequenceFeature("CDS", "cds", 4, 8, 0f, - null)); - dna2.addSequenceFeature(new SequenceFeature("CDS", "cds", 12, 12, 0f, - null)); - dna2.addSequenceFeature(new SequenceFeature("CDS", "cds", 16, 18, 0f, - null)); + + AlignmentI dna = new Alignment(new SequenceI[] { dna1, dna2, dna3 }); + dna.setDataset(null); - List mappings = new ArrayList(); MapList map = new MapList(new int[] { 4, 12, 16, 18 }, new int[] { 1, 4 }, 3, 1); AlignedCodonFrame acf = new AlignedCodonFrame(); acf.addMap(dna1.getDatasetSequence(), pep1.getDatasetSequence(), map); - mappings.add(acf); + dna.addCodonFrame(acf); map = new MapList(new int[] { 4, 8, 12, 12, 16, 18 }, new int[] { 1, 3 }, 3, 1); acf = new AlignedCodonFrame(); acf.addMap(dna2.getDatasetSequence(), pep2.getDatasetSequence(), map); - mappings.add(acf); + dna.addCodonFrame(acf); AlignmentI cds = AlignmentUtils.makeCdsAlignment(new SequenceI[] { - dna1, dna2, dna3 }, mappings, '-'); - assertEquals(2, cds.getSequences().size()); - assertEquals("GGGCCCTTTGGG", cds.getSequenceAt(0).getSequenceAsString()); - assertEquals("GGGCC---TGGG", cds.getSequenceAt(1).getSequenceAsString()); + dna1, dna2, dna3 }, dna.getDataset(), null); + List cdsSeqs = cds.getSequences(); + assertEquals(2, cdsSeqs.size()); + assertEquals("GGGCCCTTTGGG", cdsSeqs.get(0).getSequenceAsString()); + assertEquals("GGGCCTGGG", cdsSeqs.get(1).getSequenceAsString()); /* - * Verify updated mappings + * verify shared, extended alignment dataset + */ + assertSame(dna.getDataset(), cds.getDataset()); + assertTrue(dna.getDataset().getSequences() + .contains(cdsSeqs.get(0).getDatasetSequence())); + assertTrue(dna.getDataset().getSequences() + .contains(cdsSeqs.get(1).getDatasetSequence())); + + /* + * Verify 6 mappings: dna1 to cds1, cds1 to pep1, dna1 to pep1 + * and the same for dna2/cds2/pep2 */ - assertEquals(2, mappings.size()); + List mappings = cds.getCodonFrames(); + assertEquals(6, mappings.size()); /* - * Mapping from pep1 to GGGTTT in first new CDS sequence + * 2 mappings involve pep1 */ - List pep1Mapping = MappingUtils + List pep1Mappings = MappingUtils .findMappingsForSequence(pep1, mappings); - assertEquals(1, pep1Mapping.size()); + assertEquals(2, pep1Mappings.size()); + /* + * Get mapping of pep1 to cds1 and verify it * maps GPFG to 1-3,4-6,7-9,10-12 */ - SearchResults sr = MappingUtils.buildSearchResults(pep1, 1, mappings); + List pep1CdsMappings = MappingUtils + .findMappingsForSequence(cds.getSequenceAt(0), pep1Mappings); + assertEquals(1, pep1CdsMappings.size()); + SearchResults sr = MappingUtils.buildSearchResults(pep1, 1, + pep1CdsMappings); assertEquals(1, sr.getResults().size()); Match m = sr.getResults().get(0); assertEquals(cds.getSequenceAt(0).getDatasetSequence(), m.getSequence()); assertEquals(1, m.getStart()); assertEquals(3, m.getEnd()); - sr = MappingUtils.buildSearchResults(pep1, 2, mappings); + sr = MappingUtils.buildSearchResults(pep1, 2, pep1CdsMappings); m = sr.getResults().get(0); assertEquals(4, m.getStart()); assertEquals(6, m.getEnd()); - sr = MappingUtils.buildSearchResults(pep1, 3, mappings); + sr = MappingUtils.buildSearchResults(pep1, 3, pep1CdsMappings); m = sr.getResults().get(0); assertEquals(7, m.getStart()); assertEquals(9, m.getEnd()); - sr = MappingUtils.buildSearchResults(pep1, 4, mappings); + sr = MappingUtils.buildSearchResults(pep1, 4, pep1CdsMappings); m = sr.getResults().get(0); assertEquals(10, m.getStart()); assertEquals(12, m.getEnd()); /* - * GPG in pep2 map to 1-3,4-6,7-9 in second CDS sequence + * Get mapping of pep2 to cds2 and verify it + * maps GPG in pep2 to 1-3,4-6,7-9 in second CDS sequence */ - List pep2Mapping = MappingUtils + List pep2Mappings = MappingUtils .findMappingsForSequence(pep2, mappings); - assertEquals(1, pep2Mapping.size()); - sr = MappingUtils.buildSearchResults(pep2, 1, mappings); + assertEquals(2, pep2Mappings.size()); + List pep2CdsMappings = MappingUtils + .findMappingsForSequence(cds.getSequenceAt(1), pep2Mappings); + assertEquals(1, pep2CdsMappings.size()); + sr = MappingUtils.buildSearchResults(pep2, 1, pep2CdsMappings); assertEquals(1, sr.getResults().size()); m = sr.getResults().get(0); assertEquals(cds.getSequenceAt(1).getDatasetSequence(), m.getSequence()); assertEquals(1, m.getStart()); assertEquals(3, m.getEnd()); - sr = MappingUtils.buildSearchResults(pep2, 2, mappings); + sr = MappingUtils.buildSearchResults(pep2, 2, pep2CdsMappings); m = sr.getResults().get(0); assertEquals(4, m.getStart()); assertEquals(6, m.getEnd()); - sr = MappingUtils.buildSearchResults(pep2, 3, mappings); + sr = MappingUtils.buildSearchResults(pep2, 3, pep2CdsMappings); m = sr.getResults().get(0); assertEquals(7, m.getStart()); assertEquals(9, m.getEnd()); } /** - * Tests for gapped column in sequences - */ - @Test(groups = { "Functional" }) - public void testIsGappedColumn() - { - SequenceI seq1 = new Sequence("Seq1", "a--c.tc-a-g"); - SequenceI seq2 = new Sequence("Seq2", "aa---t--a-g"); - SequenceI seq3 = new Sequence("Seq3", "ag-c t-g-"); - List seqs = Arrays - .asList(new SequenceI[] { seq1, seq2, seq3 }); - // the column number is base 1 - assertFalse(AlignmentUtils.isGappedColumn(seqs, 1)); - assertFalse(AlignmentUtils.isGappedColumn(seqs, 2)); - assertTrue(AlignmentUtils.isGappedColumn(seqs, 3)); - assertFalse(AlignmentUtils.isGappedColumn(seqs, 4)); - assertTrue(AlignmentUtils.isGappedColumn(seqs, 5)); - assertFalse(AlignmentUtils.isGappedColumn(seqs, 6)); - assertFalse(AlignmentUtils.isGappedColumn(seqs, 7)); - assertFalse(AlignmentUtils.isGappedColumn(seqs, 8)); - assertFalse(AlignmentUtils.isGappedColumn(seqs, 9)); - assertTrue(AlignmentUtils.isGappedColumn(seqs, 10)); - assertFalse(AlignmentUtils.isGappedColumn(seqs, 11)); - // out of bounds: - assertTrue(AlignmentUtils.isGappedColumn(seqs, 0)); - assertTrue(AlignmentUtils.isGappedColumn(seqs, 100)); - assertTrue(AlignmentUtils.isGappedColumn(seqs, -100)); - assertTrue(AlignmentUtils.isGappedColumn(null, 0)); - } - - @Test(groups = { "Functional" }) - public void testFindCdsColumns() - { - // TODO target method belongs in a general-purpose alignment - // analysis method to find columns for feature - - /* - * NB this method assumes CDS ranges are contiguous (no introns) - */ - SequenceI gene = new Sequence("gene", "aaacccgggtttaaacccgggttt"); - SequenceI seq1 = new Sequence("Seq1", "--ac-cgGG-GGaaACC--GGtt-"); - SequenceI seq2 = new Sequence("Seq2", "AA--CCGG--g-AAA--cG-GTTt"); - seq1.createDatasetSequence(); - seq2.createDatasetSequence(); - seq1.addSequenceFeature(new SequenceFeature("CDS", "cds", 5, 6, 0f, - null)); - seq1.addSequenceFeature(new SequenceFeature("CDS", "cds", 7, 8, 0f, - null)); - seq1.addSequenceFeature(new SequenceFeature("CDS", "cds", 11, 13, 0f, - null)); - seq1.addSequenceFeature(new SequenceFeature("CDS", "cds", 14, 15, 0f, - null)); - seq2.addSequenceFeature(new SequenceFeature("CDS", "cds", 1, 2, 0f, - null)); - seq2.addSequenceFeature(new SequenceFeature("CDS", "cds", 3, 6, 0f, - null)); - seq2.addSequenceFeature(new SequenceFeature("CDS", "cds", 8, 10, 0f, - null)); - seq2.addSequenceFeature(new SequenceFeature("CDS", "cds", 12, 12, 0f, - null)); - seq2.addSequenceFeature(new SequenceFeature("CDS", "cds", 13, 15, 0f, - null)); - - List cdsColumns = AlignmentUtils.findCdsColumns(new SequenceI[] { - seq1, seq2 }); - assertEquals(4, cdsColumns.size()); - assertEquals("[1, 2]", Arrays.toString(cdsColumns.get(0))); - assertEquals("[5, 9]", Arrays.toString(cdsColumns.get(1))); - assertEquals("[11, 17]", Arrays.toString(cdsColumns.get(2))); - assertEquals("[19, 23]", Arrays.toString(cdsColumns.get(3))); - } - - /** * Test the method that realigns protein to match mapped codon alignment. */ @Test(groups = { "Functional" }) @@ -1747,4 +1720,760 @@ public class AlignmentUtilsTests assertEquals("--N-G", prot2.getSequenceAsString()); assertEquals("---XG", prot3.getSequenceAsString()); } + + /** + * Tests for the method that maps the subset of a dna sequence that has CDS + * (or subtype) feature - case where the start codon is incomplete. + */ + @Test(groups = "Functional") + public void testFindCdsPositions_fivePrimeIncomplete() + { + SequenceI dnaSeq = new Sequence("dna", "aaagGGCCCaaaTTTttt"); + dnaSeq.createDatasetSequence(); + SequenceI ds = dnaSeq.getDatasetSequence(); + + // CDS for dna 5-6 (incomplete codon), 7-9 + SequenceFeature sf = new SequenceFeature("CDS", "", 5, 9, 0f, null); + sf.setPhase("2"); // skip 2 bases to start of next codon + ds.addSequenceFeature(sf); + // CDS for dna 13-15 + sf = new SequenceFeature("CDS_predicted", "", 13, 15, 0f, null); + ds.addSequenceFeature(sf); + + List ranges = AlignmentUtils.findCdsPositions(dnaSeq); + + /* + * check the mapping starts with the first complete codon + */ + assertEquals(6, MappingUtils.getLength(ranges)); + assertEquals(2, ranges.size()); + assertEquals(7, ranges.get(0)[0]); + assertEquals(9, ranges.get(0)[1]); + assertEquals(13, ranges.get(1)[0]); + assertEquals(15, ranges.get(1)[1]); + } + + /** + * Tests for the method that maps the subset of a dna sequence that has CDS + * (or subtype) feature. + */ + @Test(groups = "Functional") + public void testFindCdsPositions() + { + SequenceI dnaSeq = new Sequence("dna", "aaaGGGcccAAATTTttt"); + dnaSeq.createDatasetSequence(); + SequenceI ds = dnaSeq.getDatasetSequence(); + + // CDS for dna 10-12 + SequenceFeature sf = new SequenceFeature("CDS_predicted", "", 10, 12, + 0f, null); + sf.setStrand("+"); + ds.addSequenceFeature(sf); + // CDS for dna 4-6 + sf = new SequenceFeature("CDS", "", 4, 6, 0f, null); + sf.setStrand("+"); + ds.addSequenceFeature(sf); + // exon feature should be ignored here + sf = new SequenceFeature("exon", "", 7, 9, 0f, null); + ds.addSequenceFeature(sf); + + List ranges = AlignmentUtils.findCdsPositions(dnaSeq); + /* + * verify ranges { [4-6], [12-10] } + * note CDS ranges are ordered ascending even if the CDS + * features are not + */ + assertEquals(6, MappingUtils.getLength(ranges)); + assertEquals(2, ranges.size()); + assertEquals(4, ranges.get(0)[0]); + assertEquals(6, ranges.get(0)[1]); + assertEquals(10, ranges.get(1)[0]); + assertEquals(12, ranges.get(1)[1]); + } + + /** + * Test the method that computes a map of codon variants for each protein + * position from "sequence_variant" features on dna + */ + @Test(groups = "Functional") + public void testBuildDnaVariantsMap() + { + SequenceI dna = new Sequence("dna", "atgAAATTTGGGCCCtag"); + MapList map = new MapList(new int[] { 1, 18 }, new int[] { 1, 5 }, 3, 1); + + /* + * first with no variants on dna + */ + LinkedHashMap[]> variantsMap = AlignmentUtils + .buildDnaVariantsMap(dna, map); + assertTrue(variantsMap.isEmpty()); + + /* + * single allele codon 1, on base 1 + */ + SequenceFeature sf1 = new SequenceFeature("sequence_variant", "", 1, 1, + 0f, null); + sf1.setValue("alleles", "T"); + sf1.setValue("ID", "sequence_variant:rs758803211"); + dna.addSequenceFeature(sf1); + + /* + * two alleles codon 2, on bases 2 and 3 (distinct variants) + */ + SequenceFeature sf2 = new SequenceFeature("sequence_variant", "", 5, 5, + 0f, null); + sf2.setValue("alleles", "T"); + sf2.setValue("ID", "sequence_variant:rs758803212"); + dna.addSequenceFeature(sf2); + SequenceFeature sf3 = new SequenceFeature("sequence_variant", "", 6, 6, + 0f, null); + sf3.setValue("alleles", "G"); + sf3.setValue("ID", "sequence_variant:rs758803213"); + dna.addSequenceFeature(sf3); + + /* + * two alleles codon 3, both on base 2 (one variant) + */ + SequenceFeature sf4 = new SequenceFeature("sequence_variant", "", 8, 8, + 0f, null); + sf4.setValue("alleles", "C, G"); + sf4.setValue("ID", "sequence_variant:rs758803214"); + dna.addSequenceFeature(sf4); + + // no alleles on codon 4 + + /* + * alleles on codon 5 on all 3 bases (distinct variants) + */ + SequenceFeature sf5 = new SequenceFeature("sequence_variant", "", 13, + 13, 0f, null); + sf5.setValue("alleles", "C, G"); // (C duplicates given base value) + sf5.setValue("ID", "sequence_variant:rs758803215"); + dna.addSequenceFeature(sf5); + SequenceFeature sf6 = new SequenceFeature("sequence_variant", "", 14, + 14, 0f, null); + sf6.setValue("alleles", "g, a"); // should force to upper-case + sf6.setValue("ID", "sequence_variant:rs758803216"); + dna.addSequenceFeature(sf6); + SequenceFeature sf7 = new SequenceFeature("sequence_variant", "", 15, + 15, 0f, null); + sf7.setValue("alleles", "A, T"); + sf7.setValue("ID", "sequence_variant:rs758803217"); + dna.addSequenceFeature(sf7); + + /* + * build map - expect variants on positions 1, 2, 3, 5 + */ + variantsMap = AlignmentUtils.buildDnaVariantsMap(dna, map); + assertEquals(4, variantsMap.size()); + + /* + * protein residue 1: variant on codon (ATG) base 1, not on 2 or 3 + */ + List[] pep1Variants = variantsMap.get(1); + assertEquals(3, pep1Variants.length); + assertEquals(1, pep1Variants[0].size()); + assertEquals("A", pep1Variants[0].get(0).base); // codon[1] base + assertSame(sf1, pep1Variants[0].get(0).variant); // codon[1] variant + assertEquals(1, pep1Variants[1].size()); + assertEquals("T", pep1Variants[1].get(0).base); // codon[2] base + assertNull(pep1Variants[1].get(0).variant); // no variant here + assertEquals(1, pep1Variants[2].size()); + assertEquals("G", pep1Variants[2].get(0).base); // codon[3] base + assertNull(pep1Variants[2].get(0).variant); // no variant here + + /* + * protein residue 2: variants on codon (AAA) bases 2 and 3 + */ + List[] pep2Variants = variantsMap.get(2); + assertEquals(3, pep2Variants.length); + assertEquals(1, pep2Variants[0].size()); + // codon[1] base recorded while processing variant on codon[2] + assertEquals("A", pep2Variants[0].get(0).base); + assertNull(pep2Variants[0].get(0).variant); // no variant here + // codon[2] base and variant: + assertEquals(1, pep2Variants[1].size()); + assertEquals("A", pep2Variants[1].get(0).base); + assertSame(sf2, pep2Variants[1].get(0).variant); + // codon[3] base was recorded when processing codon[2] variant + // and then the variant for codon[3] added to it + assertEquals(1, pep2Variants[2].size()); + assertEquals("A", pep2Variants[2].get(0).base); + assertSame(sf3, pep2Variants[2].get(0).variant); + + /* + * protein residue 3: variants on codon (TTT) base 2 only + */ + List[] pep3Variants = variantsMap.get(3); + assertEquals(3, pep3Variants.length); + assertEquals(1, pep3Variants[0].size()); + assertEquals("T", pep3Variants[0].get(0).base); // codon[1] base + assertNull(pep3Variants[0].get(0).variant); // no variant here + assertEquals(1, pep3Variants[1].size()); + assertEquals("T", pep3Variants[1].get(0).base); // codon[2] base + assertSame(sf4, pep3Variants[1].get(0).variant); // codon[2] variant + assertEquals(1, pep3Variants[2].size()); + assertEquals("T", pep3Variants[2].get(0).base); // codon[3] base + assertNull(pep3Variants[2].get(0).variant); // no variant here + + /* + * three variants on protein position 5 + */ + List[] pep5Variants = variantsMap.get(5); + assertEquals(3, pep5Variants.length); + assertEquals(1, pep5Variants[0].size()); + assertEquals("C", pep5Variants[0].get(0).base); // codon[1] base + assertSame(sf5, pep5Variants[0].get(0).variant); // codon[1] variant + assertEquals(1, pep5Variants[1].size()); + assertEquals("C", pep5Variants[1].get(0).base); // codon[2] base + assertSame(sf6, pep5Variants[1].get(0).variant); // codon[2] variant + assertEquals(1, pep5Variants[2].size()); + assertEquals("C", pep5Variants[2].get(0).base); // codon[3] base + assertSame(sf7, pep5Variants[2].get(0).variant); // codon[3] variant + } + + /** + * Tests for the method that computes all peptide variants given codon + * variants + */ + @Test(groups = "Functional") + public void testComputePeptideVariants() + { + /* + * scenario: AAATTTCCC codes for KFP, with variants + * GAA -> E + * CAA -> Q + * AAG synonymous + * AAT -> N + * TTC synonymous + * CAC,CGC -> H,R (as one variant) + */ + SequenceI peptide = new Sequence("pep/10-12", "KFP"); + + /* + * two distinct variants for codon 1 position 1 + * second one has clinical significance + */ + SequenceFeature sf1 = new SequenceFeature("sequence_variant", "", 1, 1, + 0f, null); + sf1.setValue("alleles", "A,G"); // GAA -> E + sf1.setValue("ID", "var1.125A>G"); + SequenceFeature sf2 = new SequenceFeature("sequence_variant", "", 1, 1, + 0f, null); + sf2.setValue("alleles", "A,C"); // CAA -> Q + sf2.setValue("ID", "var2"); + sf2.setValue("clinical_significance", "Dodgy"); + SequenceFeature sf3 = new SequenceFeature("sequence_variant", "", 3, 3, + 0f, null); + sf3.setValue("alleles", "A,G"); // synonymous + sf3.setValue("ID", "var3"); + sf3.setValue("clinical_significance", "None"); + SequenceFeature sf4 = new SequenceFeature("sequence_variant", "", 3, 3, + 0f, null); + sf4.setValue("alleles", "A,T"); // AAT -> N + sf4.setValue("ID", "sequence_variant:var4"); // prefix gets stripped off + sf4.setValue("clinical_significance", "Benign"); + SequenceFeature sf5 = new SequenceFeature("sequence_variant", "", 6, 6, + 0f, null); + sf5.setValue("alleles", "T,C"); // synonymous + sf5.setValue("ID", "var5"); + sf5.setValue("clinical_significance", "Bad"); + SequenceFeature sf6 = new SequenceFeature("sequence_variant", "", 8, 8, + 0f, null); + sf6.setValue("alleles", "C,A,G"); // CAC,CGC -> H,R + sf6.setValue("ID", "var6"); + sf6.setValue("clinical_significance", "Good"); + + List codon1Variants = new ArrayList(); + List codon2Variants = new ArrayList(); + List codon3Variants = new ArrayList(); + List codonVariants[] = new ArrayList[3]; + codonVariants[0] = codon1Variants; + codonVariants[1] = codon2Variants; + codonVariants[2] = codon3Variants; + + /* + * compute variants for protein position 1 + */ + codon1Variants.add(new DnaVariant("A", sf1)); + codon1Variants.add(new DnaVariant("A", sf2)); + codon2Variants.add(new DnaVariant("A")); + codon2Variants.add(new DnaVariant("A")); + codon3Variants.add(new DnaVariant("A", sf3)); + codon3Variants.add(new DnaVariant("A", sf4)); + AlignmentUtils.computePeptideVariants(peptide, 1, codonVariants); + + /* + * compute variants for protein position 2 + */ + codon1Variants.clear(); + codon2Variants.clear(); + codon3Variants.clear(); + codon1Variants.add(new DnaVariant("T")); + codon2Variants.add(new DnaVariant("T")); + codon3Variants.add(new DnaVariant("T", sf5)); + AlignmentUtils.computePeptideVariants(peptide, 2, codonVariants); + + /* + * compute variants for protein position 3 + */ + codon1Variants.clear(); + codon2Variants.clear(); + codon3Variants.clear(); + codon1Variants.add(new DnaVariant("C")); + codon2Variants.add(new DnaVariant("C", sf6)); + codon3Variants.add(new DnaVariant("C")); + AlignmentUtils.computePeptideVariants(peptide, 3, codonVariants); + + /* + * verify added sequence features for + * var1 K -> E + * var2 K -> Q + * var4 K -> N + * var6 P -> H + * var6 P -> R + */ + SequenceFeature[] sfs = peptide.getSequenceFeatures(); + assertEquals(5, sfs.length); + SequenceFeature sf = sfs[0]; + assertEquals(1, sf.getBegin()); + assertEquals(1, sf.getEnd()); + assertEquals("p.Lys1Glu", sf.getDescription()); + assertEquals("var1.125A>G", sf.getValue("ID")); + assertNull(sf.getValue("clinical_significance")); + assertEquals("ID=var1.125A>G", sf.getAttributes()); + assertEquals(1, sf.links.size()); + // link to variation is urlencoded + assertEquals( + "p.Lys1Glu var1.125A>G|http://www.ensembl.org/Homo_sapiens/Variation/Summary?v=var1.125A%3EG", + sf.links.get(0)); + assertEquals("Jalview", sf.getFeatureGroup()); + sf = sfs[1]; + assertEquals(1, sf.getBegin()); + assertEquals(1, sf.getEnd()); + assertEquals("p.Lys1Gln", sf.getDescription()); + assertEquals("var2", sf.getValue("ID")); + assertEquals("Dodgy", sf.getValue("clinical_significance")); + assertEquals("ID=var2;clinical_significance=Dodgy", sf.getAttributes()); + assertEquals(1, sf.links.size()); + assertEquals( + "p.Lys1Gln var2|http://www.ensembl.org/Homo_sapiens/Variation/Summary?v=var2", + sf.links.get(0)); + assertEquals("Jalview", sf.getFeatureGroup()); + sf = sfs[2]; + assertEquals(1, sf.getBegin()); + assertEquals(1, sf.getEnd()); + assertEquals("p.Lys1Asn", sf.getDescription()); + assertEquals("var4", sf.getValue("ID")); + assertEquals("Benign", sf.getValue("clinical_significance")); + assertEquals("ID=var4;clinical_significance=Benign", sf.getAttributes()); + assertEquals(1, sf.links.size()); + assertEquals( + "p.Lys1Asn var4|http://www.ensembl.org/Homo_sapiens/Variation/Summary?v=var4", + sf.links.get(0)); + assertEquals("Jalview", sf.getFeatureGroup()); + sf = sfs[3]; + assertEquals(3, sf.getBegin()); + assertEquals(3, sf.getEnd()); + assertEquals("p.Pro3His", sf.getDescription()); + assertEquals("var6", sf.getValue("ID")); + assertEquals("Good", sf.getValue("clinical_significance")); + assertEquals("ID=var6;clinical_significance=Good", sf.getAttributes()); + assertEquals(1, sf.links.size()); + assertEquals( + "p.Pro3His var6|http://www.ensembl.org/Homo_sapiens/Variation/Summary?v=var6", + sf.links.get(0)); + // var5 generates two distinct protein variant features + assertEquals("Jalview", sf.getFeatureGroup()); + sf = sfs[4]; + assertEquals(3, sf.getBegin()); + assertEquals(3, sf.getEnd()); + assertEquals("p.Pro3Arg", sf.getDescription()); + assertEquals("var6", sf.getValue("ID")); + assertEquals("Good", sf.getValue("clinical_significance")); + assertEquals("ID=var6;clinical_significance=Good", sf.getAttributes()); + assertEquals(1, sf.links.size()); + assertEquals( + "p.Pro3Arg var6|http://www.ensembl.org/Homo_sapiens/Variation/Summary?v=var6", + sf.links.get(0)); + assertEquals("Jalview", sf.getFeatureGroup()); + } + + /** + * Tests for the method that maps the subset of a dna sequence that has CDS + * (or subtype) feature, with CDS strand = '-' (reverse) + */ + // test turned off as currently findCdsPositions is not strand-dependent + // left in case it comes around again... + @Test(groups = "Functional", enabled = false) + public void testFindCdsPositions_reverseStrand() + { + SequenceI dnaSeq = new Sequence("dna", "aaaGGGcccAAATTTttt"); + dnaSeq.createDatasetSequence(); + SequenceI ds = dnaSeq.getDatasetSequence(); + + // CDS for dna 4-6 + SequenceFeature sf = new SequenceFeature("CDS", "", 4, 6, 0f, null); + sf.setStrand("-"); + ds.addSequenceFeature(sf); + // exon feature should be ignored here + sf = new SequenceFeature("exon", "", 7, 9, 0f, null); + ds.addSequenceFeature(sf); + // CDS for dna 10-12 + sf = new SequenceFeature("CDS_predicted", "", 10, 12, 0f, null); + sf.setStrand("-"); + ds.addSequenceFeature(sf); + + List ranges = AlignmentUtils.findCdsPositions(dnaSeq); + /* + * verify ranges { [12-10], [6-4] } + */ + assertEquals(6, MappingUtils.getLength(ranges)); + assertEquals(2, ranges.size()); + assertEquals(12, ranges.get(0)[0]); + assertEquals(10, ranges.get(0)[1]); + assertEquals(6, ranges.get(1)[0]); + assertEquals(4, ranges.get(1)[1]); + } + + /** + * Tests for the method that maps the subset of a dna sequence that has CDS + * (or subtype) feature - reverse strand case where the start codon is + * incomplete. + */ + @Test(groups = "Functional", enabled = false) + // test turned off as currently findCdsPositions is not strand-dependent + // left in case it comes around again... + public void testFindCdsPositions_reverseStrandThreePrimeIncomplete() + { + SequenceI dnaSeq = new Sequence("dna", "aaagGGCCCaaaTTTttt"); + dnaSeq.createDatasetSequence(); + SequenceI ds = dnaSeq.getDatasetSequence(); + + // CDS for dna 5-9 + SequenceFeature sf = new SequenceFeature("CDS", "", 5, 9, 0f, null); + sf.setStrand("-"); + ds.addSequenceFeature(sf); + // CDS for dna 13-15 + sf = new SequenceFeature("CDS_predicted", "", 13, 15, 0f, null); + sf.setStrand("-"); + sf.setPhase("2"); // skip 2 bases to start of next codon + ds.addSequenceFeature(sf); + + List ranges = AlignmentUtils.findCdsPositions(dnaSeq); + + /* + * check the mapping starts with the first complete codon + * expect ranges [13, 13], [9, 5] + */ + assertEquals(6, MappingUtils.getLength(ranges)); + assertEquals(2, ranges.size()); + assertEquals(13, ranges.get(0)[0]); + assertEquals(13, ranges.get(0)[1]); + assertEquals(9, ranges.get(1)[0]); + assertEquals(5, ranges.get(1)[1]); + } + + @Test(groups = "Functional") + public void testAlignAs_alternateTranscriptsUngapped() + { + SequenceI dna1 = new Sequence("dna1", "cccGGGTTTaaa"); + SequenceI dna2 = new Sequence("dna2", "CCCgggtttAAA"); + AlignmentI dna = new Alignment(new SequenceI[] { dna1, dna2 }); + ((Alignment) dna).createDatasetAlignment(); + SequenceI cds1 = new Sequence("cds1", "GGGTTT"); + SequenceI cds2 = new Sequence("cds2", "CCCAAA"); + AlignmentI cds = new Alignment(new SequenceI[] { cds1, cds2 }); + ((Alignment) cds).createDatasetAlignment(); + + AlignedCodonFrame acf = new AlignedCodonFrame(); + MapList map = new MapList(new int[] { 4, 9 }, new int[] { 1, 6 }, 1, 1); + acf.addMap(dna1.getDatasetSequence(), cds1.getDatasetSequence(), map); + map = new MapList(new int[] { 1, 3, 10, 12 }, new int[] { 1, 6 }, 1, 1); + acf.addMap(dna2.getDatasetSequence(), cds2.getDatasetSequence(), map); + + /* + * verify CDS alignment is as: + * cccGGGTTTaaa (cdna) + * CCCgggtttAAA (cdna) + * + * ---GGGTTT--- (cds) + * CCC------AAA (cds) + */ + dna.addCodonFrame(acf); + AlignmentUtils.alignAs(cds, dna); + assertEquals("---GGGTTT", cds.getSequenceAt(0).getSequenceAsString()); + assertEquals("CCC------AAA", cds.getSequenceAt(1).getSequenceAsString()); + } + + @Test(groups = { "Functional" }) + public void testAddMappedPositions() + { + SequenceI from = new Sequence("dna", "ggAA-ATcc-TT-g"); + SequenceI seq1 = new Sequence("cds", "AAATTT"); + from.createDatasetSequence(); + seq1.createDatasetSequence(); + Mapping mapping = new Mapping(seq1, new MapList( + new int[] { 3, 6, 9, 10 }, + new int[] { 1, 6 }, 1, 1)); + Map> map = new TreeMap>(); + AlignmentUtils.addMappedPositions(seq1, from, mapping, map); + + /* + * verify map has seq1 residues in columns 3,4,6,7,11,12 + */ + assertEquals(6, map.size()); + assertEquals('A', map.get(3).get(seq1).charValue()); + assertEquals('A', map.get(4).get(seq1).charValue()); + assertEquals('A', map.get(6).get(seq1).charValue()); + assertEquals('T', map.get(7).get(seq1).charValue()); + assertEquals('T', map.get(11).get(seq1).charValue()); + assertEquals('T', map.get(12).get(seq1).charValue()); + + /* + * + */ + } + + /** + * Test case where the mapping 'from' range includes a stop codon which is + * absent in the 'to' range + */ + @Test(groups = { "Functional" }) + public void testAddMappedPositions_withStopCodon() + { + SequenceI from = new Sequence("dna", "ggAA-ATcc-TT-g"); + SequenceI seq1 = new Sequence("cds", "AAATTT"); + from.createDatasetSequence(); + seq1.createDatasetSequence(); + Mapping mapping = new Mapping(seq1, new MapList( + new int[] { 3, 6, 9, 10 }, + new int[] { 1, 6 }, 1, 1)); + Map> map = new TreeMap>(); + AlignmentUtils.addMappedPositions(seq1, from, mapping, map); + + /* + * verify map has seq1 residues in columns 3,4,6,7,11,12 + */ + assertEquals(6, map.size()); + assertEquals('A', map.get(3).get(seq1).charValue()); + assertEquals('A', map.get(4).get(seq1).charValue()); + assertEquals('A', map.get(6).get(seq1).charValue()); + assertEquals('T', map.get(7).get(seq1).charValue()); + assertEquals('T', map.get(11).get(seq1).charValue()); + assertEquals('T', map.get(12).get(seq1).charValue()); + } + + /** + * Test for the case where the products for which we want CDS are specified. + * This is to represent the case where EMBL has CDS mappings to both Uniprot + * and EMBLCDSPROTEIN. makeCdsAlignment() should only return the mappings for + * the protein sequences specified. + */ + @Test(groups = { "Functional" }) + public void testMakeCdsAlignment_filterProducts() + { + SequenceI dna1 = new Sequence("dna1", "aaaGGGcccTTTaaa"); + SequenceI dna2 = new Sequence("dna2", "GGGcccTTTaaaCCC"); + SequenceI pep1 = new Sequence("Uniprot|pep1", "GF"); + SequenceI pep2 = new Sequence("Uniprot|pep2", "GFP"); + SequenceI pep3 = new Sequence("EMBL|pep3", "GF"); + SequenceI pep4 = new Sequence("EMBL|pep4", "GFP"); + dna1.createDatasetSequence(); + dna2.createDatasetSequence(); + pep1.createDatasetSequence(); + pep2.createDatasetSequence(); + pep3.createDatasetSequence(); + pep4.createDatasetSequence(); + AlignmentI dna = new Alignment(new SequenceI[] { dna1, dna2 }); + dna.setDataset(null); + AlignmentI emblPeptides = new Alignment(new SequenceI[] { pep3, pep4 }); + emblPeptides.setDataset(null); + + AlignedCodonFrame acf = new AlignedCodonFrame(); + MapList map = new MapList(new int[] { 4, 6, 10, 12 }, + new int[] { 1, 2 }, 3, 1); + acf.addMap(dna1.getDatasetSequence(), pep1.getDatasetSequence(), map); + acf.addMap(dna1.getDatasetSequence(), pep3.getDatasetSequence(), map); + dna.addCodonFrame(acf); + + acf = new AlignedCodonFrame(); + map = new MapList(new int[] { 1, 3, 7, 9, 13, 15 }, new int[] { 1, 3 }, + 3, 1); + acf.addMap(dna2.getDatasetSequence(), pep2.getDatasetSequence(), map); + acf.addMap(dna2.getDatasetSequence(), pep4.getDatasetSequence(), map); + dna.addCodonFrame(acf); + + /* + * execute method under test to find CDS for EMBL peptides only + */ + AlignmentI cds = AlignmentUtils.makeCdsAlignment(new SequenceI[] { + dna1, dna2 }, dna.getDataset(), emblPeptides.getSequencesArray()); + + assertEquals(2, cds.getSequences().size()); + assertEquals("GGGTTT", cds.getSequenceAt(0).getSequenceAsString()); + assertEquals("GGGTTTCCC", cds.getSequenceAt(1).getSequenceAsString()); + + /* + * verify shared, extended alignment dataset + */ + assertSame(dna.getDataset(), cds.getDataset()); + assertTrue(dna.getDataset().getSequences() + .contains(cds.getSequenceAt(0).getDatasetSequence())); + assertTrue(dna.getDataset().getSequences() + .contains(cds.getSequenceAt(1).getDatasetSequence())); + + /* + * Verify mappings from CDS to peptide, cDNA to CDS, and cDNA to peptide + * the mappings are on the shared alignment dataset + */ + List cdsMappings = cds.getDataset().getCodonFrames(); + /* + * 6 mappings, 2*(DNA->CDS), 2*(DNA->Pep), 2*(CDS->Pep) + */ + assertEquals(6, cdsMappings.size()); + + /* + * verify that mapping sets for dna and cds alignments are different + * [not current behaviour - all mappings are on the alignment dataset] + */ + // select -> subselect type to test. + // Assert.assertNotSame(dna.getCodonFrames(), cds.getCodonFrames()); + // assertEquals(4, dna.getCodonFrames().size()); + // assertEquals(4, cds.getCodonFrames().size()); + + /* + * Two mappings involve pep3 (dna to pep3, cds to pep3) + * Mapping from pep3 to GGGTTT in first new exon sequence + */ + List pep3Mappings = MappingUtils + .findMappingsForSequence(pep3, cdsMappings); + assertEquals(2, pep3Mappings.size()); + List mappings = MappingUtils + .findMappingsForSequence(cds.getSequenceAt(0), pep3Mappings); + assertEquals(1, mappings.size()); + + // map G to GGG + SearchResults sr = MappingUtils.buildSearchResults(pep3, 1, mappings); + assertEquals(1, sr.getResults().size()); + Match m = sr.getResults().get(0); + assertSame(cds.getSequenceAt(0).getDatasetSequence(), m.getSequence()); + assertEquals(1, m.getStart()); + assertEquals(3, m.getEnd()); + // map F to TTT + sr = MappingUtils.buildSearchResults(pep3, 2, mappings); + m = sr.getResults().get(0); + assertSame(cds.getSequenceAt(0).getDatasetSequence(), m.getSequence()); + assertEquals(4, m.getStart()); + assertEquals(6, m.getEnd()); + + /* + * Two mappings involve pep4 (dna to pep4, cds to pep4) + * Verify mapping from pep4 to GGGTTTCCC in second new exon sequence + */ + List pep4Mappings = MappingUtils + .findMappingsForSequence(pep4, cdsMappings); + assertEquals(2, pep4Mappings.size()); + mappings = MappingUtils.findMappingsForSequence(cds.getSequenceAt(1), + pep4Mappings); + assertEquals(1, mappings.size()); + // map G to GGG + sr = MappingUtils.buildSearchResults(pep4, 1, mappings); + assertEquals(1, sr.getResults().size()); + m = sr.getResults().get(0); + assertSame(cds.getSequenceAt(1).getDatasetSequence(), m.getSequence()); + assertEquals(1, m.getStart()); + assertEquals(3, m.getEnd()); + // map F to TTT + sr = MappingUtils.buildSearchResults(pep4, 2, mappings); + m = sr.getResults().get(0); + assertSame(cds.getSequenceAt(1).getDatasetSequence(), m.getSequence()); + assertEquals(4, m.getStart()); + assertEquals(6, m.getEnd()); + // map P to CCC + sr = MappingUtils.buildSearchResults(pep4, 3, mappings); + m = sr.getResults().get(0); + assertSame(cds.getSequenceAt(1).getDatasetSequence(), m.getSequence()); + assertEquals(7, m.getStart()); + assertEquals(9, m.getEnd()); + } + + /** + * Test the method that just copies aligned sequences, provided all sequences + * to be aligned share the aligned sequence's dataset + */ + @Test(groups = "Functional") + public void testAlignAsSameSequences() + { + SequenceI dna1 = new Sequence("dna1", "cccGGGTTTaaa"); + SequenceI dna2 = new Sequence("dna2", "CCCgggtttAAA"); + AlignmentI al1 = new Alignment(new SequenceI[] { dna1, dna2 }); + ((Alignment) al1).createDatasetAlignment(); + + SequenceI dna3 = new Sequence(dna1); + SequenceI dna4 = new Sequence(dna2); + assertSame(dna3.getDatasetSequence(), dna1.getDatasetSequence()); + assertSame(dna4.getDatasetSequence(), dna2.getDatasetSequence()); + String seq1 = "-cc-GG-GT-TT--aaa"; + dna3.setSequence(seq1); + String seq2 = "C--C-Cgg--gtt-tAA-A-"; + dna4.setSequence(seq2); + AlignmentI al2 = new Alignment(new SequenceI[] { dna3, dna4 }); + ((Alignment) al2).createDatasetAlignment(); + + assertTrue(AlignmentUtils.alignAsSameSequences(al1, al2)); + assertEquals(seq1, al1.getSequenceAt(0).getSequenceAsString()); + assertEquals(seq2, al1.getSequenceAt(1).getSequenceAsString()); + + /* + * add another sequence to 'aligned' - should still succeed, since + * unaligned sequences still share a dataset with aligned sequences + */ + SequenceI dna5 = new Sequence("dna5", "CCCgggtttAAA"); + dna5.createDatasetSequence(); + al2.addSequence(dna5); + assertTrue(AlignmentUtils.alignAsSameSequences(al1, al2)); + assertEquals(seq1, al1.getSequenceAt(0).getSequenceAsString()); + assertEquals(seq2, al1.getSequenceAt(1).getSequenceAsString()); + + /* + * add another sequence to 'unaligned' - should fail, since now not + * all unaligned sequences share a dataset with aligned sequences + */ + SequenceI dna6 = new Sequence("dna6", "CCCgggtttAAA"); + dna6.createDatasetSequence(); + al1.addSequence(dna6); + // JAL-2110 JBP Comment: what's the use case for this behaviour ? + assertFalse(AlignmentUtils.alignAsSameSequences(al1, al2)); + } + + @Test(groups = "Functional") + public void testAlignAsSameSequencesMultipleSubSeq() + { + SequenceI dna1 = new Sequence("dna1", "cccGGGTTTaaa"); + SequenceI dna2 = new Sequence("dna2", "CCCgggtttAAA"); + SequenceI as1 = dna1.deriveSequence(), as2 = dna1.deriveSequence() + .getSubSequence(3, 7), as3 = dna2.deriveSequence(); + as1.insertCharAt(6, 5, '-'); + String s_as1 = as1.getSequenceAsString(); + as2.insertCharAt(6, 5, '-'); + String s_as2 = as2.getSequenceAsString(); + as3.insertCharAt(6, 5, '-'); + String s_as3 = as3.getSequenceAsString(); + AlignmentI aligned = new Alignment(new SequenceI[] { as1, as2, as3 }); + + // why do we need to cast this still ? + ((Alignment) aligned).createDatasetAlignment(); + SequenceI uas1 = dna1.deriveSequence(), uas2 = dna1.deriveSequence() + .getSubSequence(3, 7), uas3 = dna2.deriveSequence(); + AlignmentI tobealigned = new Alignment(new SequenceI[] { uas1, uas2, + uas3 }); + ((Alignment) tobealigned).createDatasetAlignment(); + + assertTrue(AlignmentUtils.alignAsSameSequences(tobealigned, aligned)); + assertEquals(s_as1, uas1.getSequenceAsString()); + assertEquals(s_as2, uas2.getSequenceAsString()); + assertEquals(s_as3, uas3.getSequenceAsString()); + } + }