X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=test%2Fjalview%2Fanalysis%2FAlignmentUtilsTests.java;h=8bdd7403bf34d573b8052608d20dbd85019cc106;hb=547575f4b08f2cc25adcd438e8bce24e4e7af04e;hp=2104d4a404d8fa6af7c2be4a194edbc9ee278846;hpb=efde43fab576590026bbe0ced39e13426aa6532f;p=jalview.git diff --git a/test/jalview/analysis/AlignmentUtilsTests.java b/test/jalview/analysis/AlignmentUtilsTests.java index 2104d4a..8bdd740 100644 --- a/test/jalview/analysis/AlignmentUtilsTests.java +++ b/test/jalview/analysis/AlignmentUtilsTests.java @@ -1,40 +1,2009 @@ +/* + * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$) + * Copyright (C) $$Year-Rel$$ The Jalview Authors + * + * This file is part of Jalview. + * + * Jalview is free software: you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation, either version 3 + * of the License, or (at your option) any later version. + * + * Jalview is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty + * of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with Jalview. If not, see . + * The Jalview Authors are detailed in the 'AUTHORS' file. + */ package jalview.analysis; -import static org.junit.Assert.assertTrue; - -import org.junit.Test; +import static org.testng.AssertJUnit.assertEquals; +import static org.testng.AssertJUnit.assertFalse; +import static org.testng.AssertJUnit.assertNull; +import static org.testng.AssertJUnit.assertSame; +import static org.testng.AssertJUnit.assertTrue; +import jalview.datamodel.AlignedCodonFrame; import jalview.datamodel.Alignment; +import jalview.datamodel.AlignmentAnnotation; import jalview.datamodel.AlignmentI; +import jalview.datamodel.Annotation; +import jalview.datamodel.DBRefEntry; +import jalview.datamodel.Mapping; +import jalview.datamodel.SearchResults; +import jalview.datamodel.SearchResults.Match; import jalview.datamodel.Sequence; +import jalview.datamodel.SequenceFeature; import jalview.datamodel.SequenceI; import jalview.io.AppletFormatAdapter; +import jalview.io.FormatAdapter; +import jalview.util.MapList; +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; -public class AlignmentUtilsTests +import org.testng.annotations.Test; + +public class AlignmentUtilsTests { - public static Sequence ts=new Sequence("short","ASDASDASDASDASDASDASDASDASDASDASDASDASD"); - @Test - public void testExpandFlanks() + // @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"); + + @Test(groups = { "Functional" }) + public void testExpandContext() { AlignmentI al = new Alignment(new Sequence[] {}); - for (int i=4;i<14;i+=3) + for (int i = 4; i < 14; i += 2) { - SequenceI s1=ts.deriveSequence().getSubSequence(i, i+7); + SequenceI s1 = ts.deriveSequence().getSubSequence(i, i + 7); al.addSequence(s1); } - System.out.println(new AppletFormatAdapter().formatSequences("Clustal", al, true)); - for (int flnk=-1;flnk<25; flnk++) + System.out.println(new AppletFormatAdapter().formatSequences("Clustal", + al, true)); + for (int flnk = -1; flnk < 25; flnk++) { - AlignmentI exp; - System.out.println("\nFlank size: "+flnk); - System.out.println(new AppletFormatAdapter().formatSequences("Clustal", exp=AlignmentUtils.expandContext(al, flnk), true)); - if (flnk==-1) { - for (SequenceI sq:exp.getSequences()) + AlignmentI exp = AlignmentUtils.expandContext(al, flnk); + System.out.println("\nFlank size: " + flnk); + System.out.println(new AppletFormatAdapter().formatSequences( + "Clustal", exp, true)); + if (flnk == -1) { + /* + * Full expansion to complete sequences + */ + for (SequenceI sq : exp.getSequences()) + { String ung = sq.getSequenceAsString().replaceAll("-+", ""); - assertTrue("Flanking sequence not the same as original dataset sequence.\n"+ung+"\n"+sq.getDatasetSequence().getSequenceAsString(),ung.equalsIgnoreCase(sq.getDatasetSequence().getSequenceAsString())); + final String errorMsg = "Flanking sequence not the same as original dataset sequence.\n" + + ung + + "\n" + + sq.getDatasetSequence().getSequenceAsString(); + assertTrue(errorMsg, ung.equalsIgnoreCase(sq.getDatasetSequence() + .getSequenceAsString())); + } } + else if (flnk == 24) + { + /* + * Last sequence is fully expanded, others have leading gaps to match + */ + assertTrue(exp.getSequenceAt(4).getSequenceAsString() + .startsWith("abc")); + assertTrue(exp.getSequenceAt(3).getSequenceAsString() + .startsWith("--abc")); + assertTrue(exp.getSequenceAt(2).getSequenceAsString() + .startsWith("----abc")); + assertTrue(exp.getSequenceAt(1).getSequenceAsString() + .startsWith("------abc")); + assertTrue(exp.getSequenceAt(0).getSequenceAsString() + .startsWith("--------abc")); } - } + } + } + + /** + * Test that annotations are correctly adjusted by expandContext + */ + @Test(groups = { "Functional" }) + public void testExpandContext_annotation() + { + AlignmentI al = new Alignment(new Sequence[] {}); + SequenceI ds = new Sequence("Seq1", "ABCDEFGHI"); + // subsequence DEF: + SequenceI seq1 = ds.deriveSequence().getSubSequence(3, 6); + al.addSequence(seq1); + + /* + * Annotate DEF with 4/5/6 respectively + */ + Annotation[] anns = new Annotation[] { new Annotation(4), + new Annotation(5), new Annotation(6) }; + AlignmentAnnotation ann = new AlignmentAnnotation("SS", + "secondary structure", anns); + seq1.addAlignmentAnnotation(ann); + + /* + * The annotations array should match aligned positions + */ + assertEquals(3, ann.annotations.length); + assertEquals(4, ann.annotations[0].value, 0.001); + assertEquals(5, ann.annotations[1].value, 0.001); + assertEquals(6, ann.annotations[2].value, 0.001); + + /* + * Check annotation to sequence position mappings before expanding the + * sequence; these are set up in Sequence.addAlignmentAnnotation -> + * Annotation.setSequenceRef -> createSequenceMappings + */ + assertNull(ann.getAnnotationForPosition(1)); + assertNull(ann.getAnnotationForPosition(2)); + assertNull(ann.getAnnotationForPosition(3)); + assertEquals(4, ann.getAnnotationForPosition(4).value, 0.001); + assertEquals(5, ann.getAnnotationForPosition(5).value, 0.001); + assertEquals(6, ann.getAnnotationForPosition(6).value, 0.001); + assertNull(ann.getAnnotationForPosition(7)); + assertNull(ann.getAnnotationForPosition(8)); + assertNull(ann.getAnnotationForPosition(9)); + + /* + * Expand the subsequence to the full sequence abcDEFghi + */ + AlignmentI expanded = AlignmentUtils.expandContext(al, -1); + assertEquals("abcDEFghi", expanded.getSequenceAt(0) + .getSequenceAsString()); + + /* + * Confirm the alignment and sequence have the same SS annotation, + * referencing the expanded sequence + */ + ann = expanded.getSequenceAt(0).getAnnotation()[0]; + assertSame(ann, expanded.getAlignmentAnnotation()[0]); + assertSame(expanded.getSequenceAt(0), ann.sequenceRef); + + /* + * The annotations array should have null values except for annotated + * positions + */ + assertNull(ann.annotations[0]); + assertNull(ann.annotations[1]); + assertNull(ann.annotations[2]); + assertEquals(4, ann.annotations[3].value, 0.001); + assertEquals(5, ann.annotations[4].value, 0.001); + assertEquals(6, ann.annotations[5].value, 0.001); + assertNull(ann.annotations[6]); + assertNull(ann.annotations[7]); + assertNull(ann.annotations[8]); + + /* + * sequence position mappings should be unchanged + */ + assertNull(ann.getAnnotationForPosition(1)); + assertNull(ann.getAnnotationForPosition(2)); + assertNull(ann.getAnnotationForPosition(3)); + assertEquals(4, ann.getAnnotationForPosition(4).value, 0.001); + assertEquals(5, ann.getAnnotationForPosition(5).value, 0.001); + assertEquals(6, ann.getAnnotationForPosition(6).value, 0.001); + assertNull(ann.getAnnotationForPosition(7)); + assertNull(ann.getAnnotationForPosition(8)); + assertNull(ann.getAnnotationForPosition(9)); + } + + /** + * Test method that returns a map of lists of sequences by sequence name. + * + * @throws IOException + */ + @Test(groups = { "Functional" }) + public void testGetSequencesByName() throws IOException + { + final String data = ">Seq1Name\nKQYL\n" + ">Seq2Name\nRFPW\n" + + ">Seq1Name\nABCD\n"; + AlignmentI al = loadAlignment(data, "FASTA"); + Map> map = AlignmentUtils + .getSequencesByName(al); + assertEquals(2, map.keySet().size()); + assertEquals(2, map.get("Seq1Name").size()); + assertEquals("KQYL", map.get("Seq1Name").get(0).getSequenceAsString()); + assertEquals("ABCD", map.get("Seq1Name").get(1).getSequenceAsString()); + assertEquals(1, map.get("Seq2Name").size()); + assertEquals("RFPW", map.get("Seq2Name").get(0).getSequenceAsString()); + } + + /** + * Helper method to load an alignment and ensure dataset sequences are set up. + * + * @param data + * @param format + * TODO + * @return + * @throws IOException + */ + protected AlignmentI loadAlignment(final String data, String format) + throws IOException + { + AlignmentI a = new FormatAdapter().readFile(data, + AppletFormatAdapter.PASTE, format); + a.setDataset(null); + return a; + } + + /** + * Test mapping of protein to cDNA, for the case where we have no sequence + * cross-references, so mappings are made first-served 1-1 where sequences + * translate. + * + * @throws IOException + */ + @Test(groups = { "Functional" }) + public void testMapProteinAlignmentToCdna_noXrefs() throws IOException + { + List protseqs = new ArrayList(); + protseqs.add(new Sequence("UNIPROT|V12345", "EIQ")); + protseqs.add(new Sequence("UNIPROT|V12346", "EIQ")); + protseqs.add(new Sequence("UNIPROT|V12347", "SAR")); + AlignmentI protein = new Alignment(protseqs.toArray(new SequenceI[3])); + protein.setDataset(null); + + List dnaseqs = new ArrayList(); + dnaseqs.add(new Sequence("EMBL|A11111", "TCAGCACGC")); // = SAR + dnaseqs.add(new Sequence("EMBL|A22222", "GAGATACAA")); // = EIQ + dnaseqs.add(new Sequence("EMBL|A33333", "GAAATCCAG")); // = EIQ + dnaseqs.add(new Sequence("EMBL|A44444", "GAAATTCAG")); // = EIQ + AlignmentI cdna = new Alignment(dnaseqs.toArray(new SequenceI[4])); + cdna.setDataset(null); + + assertTrue(AlignmentUtils.mapProteinAlignmentToCdna(protein, cdna)); + + // 3 mappings made, each from 1 to 1 sequence + assertEquals(3, protein.getCodonFrames().size()); + assertEquals(1, protein.getCodonFrame(protein.getSequenceAt(0)).size()); + assertEquals(1, protein.getCodonFrame(protein.getSequenceAt(1)).size()); + assertEquals(1, protein.getCodonFrame(protein.getSequenceAt(2)).size()); + + // V12345 mapped to A22222 + AlignedCodonFrame acf = protein.getCodonFrame(protein.getSequenceAt(0)) + .get(0); + assertEquals(1, acf.getdnaSeqs().length); + assertEquals(cdna.getSequenceAt(1).getDatasetSequence(), + acf.getdnaSeqs()[0]); + Mapping[] protMappings = acf.getProtMappings(); + assertEquals(1, protMappings.length); + MapList mapList = protMappings[0].getMap(); + assertEquals(3, mapList.getFromRatio()); + assertEquals(1, mapList.getToRatio()); + assertTrue(Arrays.equals(new int[] { 1, 9 }, mapList.getFromRanges() + .get(0))); + assertEquals(1, mapList.getFromRanges().size()); + assertTrue(Arrays.equals(new int[] { 1, 3 }, + mapList.getToRanges().get(0))); + assertEquals(1, mapList.getToRanges().size()); + + // V12346 mapped to A33333 + acf = protein.getCodonFrame(protein.getSequenceAt(1)).get(0); + assertEquals(1, acf.getdnaSeqs().length); + assertEquals(cdna.getSequenceAt(2).getDatasetSequence(), + acf.getdnaSeqs()[0]); + + // V12347 mapped to A11111 + acf = protein.getCodonFrame(protein.getSequenceAt(2)).get(0); + assertEquals(1, acf.getdnaSeqs().length); + assertEquals(cdna.getSequenceAt(0).getDatasetSequence(), + acf.getdnaSeqs()[0]); + + // no mapping involving the 'extra' A44444 + assertTrue(protein.getCodonFrame(cdna.getSequenceAt(3)).isEmpty()); + } + + /** + * Test for the alignSequenceAs method that takes two sequences and a mapping. + */ + @Test(groups = { "Functional" }) + public void testAlignSequenceAs_withMapping_noIntrons() + { + MapList map = new MapList(new int[] { 1, 6 }, new int[] { 1, 2 }, 3, 1); + + /* + * No existing gaps in dna: + */ + checkAlignSequenceAs("GGGAAA", "-A-L-", false, false, map, + "---GGG---AAA"); + + /* + * Now introduce gaps in dna but ignore them when realigning. + */ + checkAlignSequenceAs("-G-G-G-A-A-A-", "-A-L-", false, false, map, + "---GGG---AAA"); + + /* + * Now include gaps in dna when realigning. First retaining 'mapped' gaps + * only, i.e. those within the exon region. + */ + checkAlignSequenceAs("-G-G--G-A--A-A-", "-A-L-", true, false, map, + "---G-G--G---A--A-A"); + + /* + * Include all gaps in dna when realigning (within and without the exon + * region). The leading gap, and the gaps between codons, are subsumed by + * the protein alignment gap. + */ + checkAlignSequenceAs("-G-GG--AA-A---", "-A-L-", true, true, map, + "---G-GG---AA-A---"); + + /* + * Include only unmapped gaps in dna when realigning (outside the exon + * region). The leading gap, and the gaps between codons, are subsumed by + * the protein alignment gap. + */ + checkAlignSequenceAs("-G-GG--AA-A-", "-A-L-", false, true, map, + "---GGG---AAA---"); + } + + /** + * Test for the alignSequenceAs method that takes two sequences and a mapping. + */ + @Test(groups = { "Functional" }) + public void testAlignSequenceAs_withMapping_withIntrons() + { + /* + * Exons at codon 2 (AAA) and 4 (TTT) + */ + MapList map = new MapList(new int[] { 4, 6, 10, 12 }, + new int[] { 1, 2 }, 3, 1); + + /* + * Simple case: no gaps in dna + */ + checkAlignSequenceAs("GGGAAACCCTTTGGG", "--A-L-", false, false, map, + "GGG---AAACCCTTTGGG"); + + /* + * Add gaps to dna - but ignore when realigning. + */ + checkAlignSequenceAs("-G-G-G--A--A---AC-CC-T-TT-GG-G-", "--A-L-", + false, false, map, "GGG---AAACCCTTTGGG"); + + /* + * Add gaps to dna - include within exons only when realigning. + */ + checkAlignSequenceAs("-G-G-G--A--A---A-C-CC-T-TT-GG-G-", "--A-L-", + true, false, map, "GGG---A--A---ACCCT-TTGGG"); + + /* + * Include gaps outside exons only when realigning. + */ + checkAlignSequenceAs("-G-G-G--A--A---A-C-CC-T-TT-GG-G-", "--A-L-", + false, true, map, "-G-G-GAAAC-CCTTT-GG-G-"); + + /* + * Include gaps following first intron if we are 'preserving mapped gaps' + */ + checkAlignSequenceAs("-G-G-G--A--A---A-C-CC-T-TT-GG-G-", "--A-L-", + true, true, map, "-G-G-G--A--A---A-C-CC-T-TT-GG-G-"); + + /* + * Include all gaps in dna when realigning. + */ + checkAlignSequenceAs("-G-G-G--A--A---A-C-CC-T-TT-GG-G-", "--A-L-", + true, true, map, "-G-G-G--A--A---A-C-CC-T-TT-GG-G-"); + } + + /** + * Test for the case where not all of the protein sequence is mapped to cDNA. + */ + @Test(groups = { "Functional" }) + public void testAlignSequenceAs_withMapping_withUnmappedProtein() + { + /* + * Exons at codon 2 (AAA) and 4 (TTT) mapped to A and P + */ + final MapList map = new MapList(new int[] { 4, 6, 10, 12 }, new int[] { + 1, 1, 3, 3 }, 3, 1); + + /* + * -L- 'aligns' ccc------ + */ + checkAlignSequenceAs("gggAAAcccTTTggg", "-A-L-P-", false, false, map, + "gggAAAccc------TTTggg"); + } + + /** + * Helper method that performs and verifies the method under test. + * + * @param alignee + * the sequence to be realigned + * @param alignModel + * the sequence whose alignment is to be copied + * @param preserveMappedGaps + * @param preserveUnmappedGaps + * @param map + * @param expected + */ + protected void checkAlignSequenceAs(final String alignee, + final String alignModel, final boolean preserveMappedGaps, + final boolean preserveUnmappedGaps, MapList map, + final String expected) + { + SequenceI alignMe = new Sequence("Seq1", alignee); + alignMe.createDatasetSequence(); + SequenceI alignFrom = new Sequence("Seq2", alignModel); + alignFrom.createDatasetSequence(); + AlignedCodonFrame acf = new AlignedCodonFrame(); + acf.addMap(alignMe.getDatasetSequence(), alignFrom.getDatasetSequence(), map); + + AlignmentUtils.alignSequenceAs(alignMe, alignFrom, acf, "---", '-', + preserveMappedGaps, preserveUnmappedGaps); + assertEquals(expected, alignMe.getSequenceAsString()); + } + + /** + * Test for the alignSequenceAs method where we preserve gaps in introns only. + */ + @Test(groups = { "Functional" }) + public void testAlignSequenceAs_keepIntronGapsOnly() + { + + /* + * Intron GGGAAA followed by exon CCCTTT + */ + MapList map = new MapList(new int[] { 7, 12 }, new int[] { 1, 2 }, 3, 1); + + checkAlignSequenceAs("GG-G-AA-A-C-CC-T-TT", "AL", false, true, map, + "GG-G-AA-ACCCTTT"); + } + + /** + * 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" }) + public void testAlignProteinAsDna() + { + // seq1 codons are [1,2,3] [4,5,6] [7,8,9] [10,11,12] + SequenceI dna1 = new Sequence("Seq1", "TGCCATTACCAG-"); + // seq2 codons are [1,3,4] [5,6,7] [8,9,10] [11,12,13] + SequenceI dna2 = new Sequence("Seq2", "T-GCCATTACCAG"); + // seq3 codons are [1,2,3] [4,5,7] [8,9,10] [11,12,13] + SequenceI dna3 = new Sequence("Seq3", "TGCCA-TTACCAG"); + AlignmentI dna = new Alignment(new SequenceI[] { dna1, dna2, dna3 }); + dna.setDataset(null); + + // protein alignment will be realigned like dna + SequenceI prot1 = new Sequence("Seq1", "CHYQ"); + SequenceI prot2 = new Sequence("Seq2", "CHYQ"); + SequenceI prot3 = new Sequence("Seq3", "CHYQ"); + SequenceI prot4 = new Sequence("Seq4", "R-QSV"); // unmapped, unchanged + AlignmentI protein = new Alignment(new SequenceI[] { prot1, prot2, + prot3, prot4 }); + protein.setDataset(null); + + MapList map = new MapList(new int[] { 1, 12 }, new int[] { 1, 4 }, 3, 1); + AlignedCodonFrame acf = new AlignedCodonFrame(); + acf.addMap(dna1.getDatasetSequence(), prot1.getDatasetSequence(), map); + acf.addMap(dna2.getDatasetSequence(), prot2.getDatasetSequence(), map); + acf.addMap(dna3.getDatasetSequence(), prot3.getDatasetSequence(), map); + ArrayList acfs = new ArrayList(); + acfs.add(acf); + protein.setCodonFrames(acfs); + + /* + * Translated codon order is [1,2,3] [1,3,4] [4,5,6] [4,5,7] [5,6,7] [7,8,9] + * [8,9,10] [10,11,12] [11,12,13] + */ + AlignmentUtils.alignProteinAsDna(protein, dna); + assertEquals("C-H--Y-Q-", prot1.getSequenceAsString()); + assertEquals("-C--H-Y-Q", prot2.getSequenceAsString()); + assertEquals("C--H--Y-Q", prot3.getSequenceAsString()); + assertEquals("R-QSV", prot4.getSequenceAsString()); + } + + /** + * Test the method that tests whether a CDNA sequence translates to a protein + * sequence + */ + @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 extra start codon (not in protein) + assertTrue(AlignmentUtils.translatesAs("atgtttcccaaaggg".toCharArray(), + 3, "FPKG".toCharArray())); + // with stop codon1 (not in protein) + assertTrue(AlignmentUtils.translatesAs("tttcccaaagggtaa".toCharArray(), + 0, "FPKG".toCharArray())); + // with stop codon1 (in protein as *) + assertTrue(AlignmentUtils.translatesAs("tttcccaaagggtaa".toCharArray(), + 0, "FPKG*".toCharArray())); + // with stop codon2 (not in protein) + assertTrue(AlignmentUtils.translatesAs("tttcccaaagggtag".toCharArray(), + 0, "FPKG".toCharArray())); + // with stop codon3 (not in protein) + assertTrue(AlignmentUtils.translatesAs("tttcccaaagggtga".toCharArray(), + 0, "FPKG".toCharArray())); + // with start and stop codon1 + assertTrue(AlignmentUtils.translatesAs( + "atgtttcccaaagggtaa".toCharArray(), 3, "FPKG".toCharArray())); + // with start and stop codon1 (in protein as *) + assertTrue(AlignmentUtils.translatesAs( + "atgtttcccaaagggtaa".toCharArray(), 3, "FPKG*".toCharArray())); + // with start and stop codon2 + assertTrue(AlignmentUtils.translatesAs( + "atgtttcccaaagggtag".toCharArray(), 3, "FPKG".toCharArray())); + // with start and stop codon3 + assertTrue(AlignmentUtils.translatesAs( + "atgtttcccaaagggtga".toCharArray(), 3, "FPKG".toCharArray())); + + // with embedded stop codons + assertTrue(AlignmentUtils.translatesAs( + "atgtttTAGcccaaaTAAgggtga".toCharArray(), 3, + "F*PK*G".toCharArray())); + + // 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())); + } + + /** + * Test mapping of protein to cDNA, for cases where the cDNA has start and/or + * stop codons in addition to the protein coding sequence. + * + * @throws IOException + */ + @Test(groups = { "Functional" }) + public void testMapProteinAlignmentToCdna_withStartAndStopCodons() + throws IOException + { + List protseqs = new ArrayList(); + protseqs.add(new Sequence("UNIPROT|V12345", "EIQ")); + protseqs.add(new Sequence("UNIPROT|V12346", "EIQ")); + protseqs.add(new Sequence("UNIPROT|V12347", "SAR")); + AlignmentI protein = new Alignment(protseqs.toArray(new SequenceI[3])); + protein.setDataset(null); + + List dnaseqs = new ArrayList(); + // start + SAR: + dnaseqs.add(new Sequence("EMBL|A11111", "ATGTCAGCACGC")); + // = EIQ + stop + dnaseqs.add(new Sequence("EMBL|A22222", "GAGATACAATAA")); + // = start +EIQ + stop + dnaseqs.add(new Sequence("EMBL|A33333", "ATGGAAATCCAGTAG")); + dnaseqs.add(new Sequence("EMBL|A44444", "GAAATTCAG")); + AlignmentI cdna = new Alignment(dnaseqs.toArray(new SequenceI[4])); + cdna.setDataset(null); + + assertTrue(AlignmentUtils.mapProteinAlignmentToCdna(protein, cdna)); + + // 3 mappings made, each from 1 to 1 sequence + assertEquals(3, protein.getCodonFrames().size()); + assertEquals(1, protein.getCodonFrame(protein.getSequenceAt(0)).size()); + assertEquals(1, protein.getCodonFrame(protein.getSequenceAt(1)).size()); + assertEquals(1, protein.getCodonFrame(protein.getSequenceAt(2)).size()); + + // V12345 mapped from A22222 + AlignedCodonFrame acf = protein.getCodonFrame(protein.getSequenceAt(0)) + .get(0); + assertEquals(1, acf.getdnaSeqs().length); + assertEquals(cdna.getSequenceAt(1).getDatasetSequence(), + acf.getdnaSeqs()[0]); + Mapping[] protMappings = acf.getProtMappings(); + assertEquals(1, protMappings.length); + MapList mapList = protMappings[0].getMap(); + assertEquals(3, mapList.getFromRatio()); + assertEquals(1, mapList.getToRatio()); + assertTrue(Arrays.equals(new int[] { 1, 9 }, mapList.getFromRanges() + .get(0))); + assertEquals(1, mapList.getFromRanges().size()); + assertTrue(Arrays.equals(new int[] { 1, 3 }, + mapList.getToRanges().get(0))); + assertEquals(1, mapList.getToRanges().size()); + + // V12346 mapped from A33333 starting position 4 + acf = protein.getCodonFrame(protein.getSequenceAt(1)).get(0); + assertEquals(1, acf.getdnaSeqs().length); + assertEquals(cdna.getSequenceAt(2).getDatasetSequence(), + acf.getdnaSeqs()[0]); + protMappings = acf.getProtMappings(); + assertEquals(1, protMappings.length); + mapList = protMappings[0].getMap(); + assertEquals(3, mapList.getFromRatio()); + assertEquals(1, mapList.getToRatio()); + assertTrue(Arrays.equals(new int[] { 4, 12 }, mapList.getFromRanges() + .get(0))); + assertEquals(1, mapList.getFromRanges().size()); + assertTrue(Arrays.equals(new int[] { 1, 3 }, + mapList.getToRanges().get(0))); + assertEquals(1, mapList.getToRanges().size()); + + // V12347 mapped to A11111 starting position 4 + acf = protein.getCodonFrame(protein.getSequenceAt(2)).get(0); + assertEquals(1, acf.getdnaSeqs().length); + assertEquals(cdna.getSequenceAt(0).getDatasetSequence(), + acf.getdnaSeqs()[0]); + protMappings = acf.getProtMappings(); + assertEquals(1, protMappings.length); + mapList = protMappings[0].getMap(); + assertEquals(3, mapList.getFromRatio()); + assertEquals(1, mapList.getToRatio()); + assertTrue(Arrays.equals(new int[] { 4, 12 }, mapList.getFromRanges() + .get(0))); + assertEquals(1, mapList.getFromRanges().size()); + assertTrue(Arrays.equals(new int[] { 1, 3 }, + mapList.getToRanges().get(0))); + assertEquals(1, mapList.getToRanges().size()); + + // no mapping involving the 'extra' A44444 + assertTrue(protein.getCodonFrame(cdna.getSequenceAt(3)).isEmpty()); + } + + /** + * Test mapping of protein to cDNA, for the case where we have some sequence + * cross-references. Verify that 1-to-many mappings are made where + * cross-references exist and sequences are mappable. + * + * @throws IOException + */ + @Test(groups = { "Functional" }) + public void testMapProteinAlignmentToCdna_withXrefs() throws IOException + { + List protseqs = new ArrayList(); + protseqs.add(new Sequence("UNIPROT|V12345", "EIQ")); + protseqs.add(new Sequence("UNIPROT|V12346", "EIQ")); + protseqs.add(new Sequence("UNIPROT|V12347", "SAR")); + AlignmentI protein = new Alignment(protseqs.toArray(new SequenceI[3])); + protein.setDataset(null); + + List dnaseqs = new ArrayList(); + dnaseqs.add(new Sequence("EMBL|A11111", "TCAGCACGC")); // = SAR + dnaseqs.add(new Sequence("EMBL|A22222", "ATGGAGATACAA")); // = start + EIQ + dnaseqs.add(new Sequence("EMBL|A33333", "GAAATCCAG")); // = EIQ + dnaseqs.add(new Sequence("EMBL|A44444", "GAAATTCAG")); // = EIQ + dnaseqs.add(new Sequence("EMBL|A55555", "GAGATTCAG")); // = EIQ + AlignmentI cdna = new Alignment(dnaseqs.toArray(new SequenceI[5])); + cdna.setDataset(null); + + // Xref A22222 to V12345 (should get mapped) + dnaseqs.get(1).addDBRef(new DBRefEntry("UNIPROT", "1", "V12345")); + // Xref V12345 to A44444 (should get mapped) + protseqs.get(0).addDBRef(new DBRefEntry("EMBL", "1", "A44444")); + // Xref A33333 to V12347 (sequence mismatch - should not get mapped) + dnaseqs.get(2).addDBRef(new DBRefEntry("UNIPROT", "1", "V12347")); + // as V12345 is mapped to A22222 and A44444, this leaves V12346 unmapped. + // it should get paired up with the unmapped A33333 + // A11111 should be mapped to V12347 + // A55555 is spare and has no xref so is not mapped + + assertTrue(AlignmentUtils.mapProteinAlignmentToCdna(protein, cdna)); + + // 4 protein mappings made for 3 proteins, 2 to V12345, 1 each to V12346/7 + assertEquals(3, protein.getCodonFrames().size()); + assertEquals(1, protein.getCodonFrame(protein.getSequenceAt(0)).size()); + assertEquals(1, protein.getCodonFrame(protein.getSequenceAt(1)).size()); + assertEquals(1, protein.getCodonFrame(protein.getSequenceAt(2)).size()); + + // one mapping for each of the first 4 cDNA sequences + assertEquals(1, protein.getCodonFrame(cdna.getSequenceAt(0)).size()); + assertEquals(1, protein.getCodonFrame(cdna.getSequenceAt(1)).size()); + assertEquals(1, protein.getCodonFrame(cdna.getSequenceAt(2)).size()); + assertEquals(1, protein.getCodonFrame(cdna.getSequenceAt(3)).size()); + + // V12345 mapped to A22222 and A44444 + AlignedCodonFrame acf = protein.getCodonFrame(protein.getSequenceAt(0)) + .get(0); + assertEquals(2, acf.getdnaSeqs().length); + assertEquals(cdna.getSequenceAt(1).getDatasetSequence(), + acf.getdnaSeqs()[0]); + assertEquals(cdna.getSequenceAt(3).getDatasetSequence(), + acf.getdnaSeqs()[1]); + + // V12346 mapped to A33333 + acf = protein.getCodonFrame(protein.getSequenceAt(1)).get(0); + assertEquals(1, acf.getdnaSeqs().length); + assertEquals(cdna.getSequenceAt(2).getDatasetSequence(), + acf.getdnaSeqs()[0]); + + // V12347 mapped to A11111 + acf = protein.getCodonFrame(protein.getSequenceAt(2)).get(0); + assertEquals(1, acf.getdnaSeqs().length); + assertEquals(cdna.getSequenceAt(0).getDatasetSequence(), + acf.getdnaSeqs()[0]); + + // no mapping involving the 'extra' A55555 + assertTrue(protein.getCodonFrame(cdna.getSequenceAt(4)).isEmpty()); + } + + /** + * Test mapping of protein to cDNA, for the case where we have some sequence + * cross-references. Verify that once we have made an xref mapping we don't + * also map un-xrefd sequeces. + * + * @throws IOException + */ + @Test(groups = { "Functional" }) + public void testMapProteinAlignmentToCdna_prioritiseXrefs() + throws IOException + { + List protseqs = new ArrayList(); + protseqs.add(new Sequence("UNIPROT|V12345", "EIQ")); + protseqs.add(new Sequence("UNIPROT|V12346", "EIQ")); + AlignmentI protein = new Alignment( + protseqs.toArray(new SequenceI[protseqs.size()])); + protein.setDataset(null); + + List dnaseqs = new ArrayList(); + dnaseqs.add(new Sequence("EMBL|A11111", "GAAATCCAG")); // = EIQ + dnaseqs.add(new Sequence("EMBL|A22222", "GAAATTCAG")); // = EIQ + AlignmentI cdna = new Alignment(dnaseqs.toArray(new SequenceI[dnaseqs + .size()])); + cdna.setDataset(null); + + // Xref A22222 to V12345 (should get mapped) + // A11111 should then be mapped to the unmapped V12346 + dnaseqs.get(1).addDBRef(new DBRefEntry("UNIPROT", "1", "V12345")); + + assertTrue(AlignmentUtils.mapProteinAlignmentToCdna(protein, cdna)); + + // 2 protein mappings made + assertEquals(2, protein.getCodonFrames().size()); + assertEquals(1, protein.getCodonFrame(protein.getSequenceAt(0)).size()); + assertEquals(1, protein.getCodonFrame(protein.getSequenceAt(1)).size()); + + // one mapping for each of the cDNA sequences + assertEquals(1, protein.getCodonFrame(cdna.getSequenceAt(0)).size()); + assertEquals(1, protein.getCodonFrame(cdna.getSequenceAt(1)).size()); + + // V12345 mapped to A22222 + AlignedCodonFrame acf = protein.getCodonFrame(protein.getSequenceAt(0)) + .get(0); + assertEquals(1, acf.getdnaSeqs().length); + assertEquals(cdna.getSequenceAt(1).getDatasetSequence(), + acf.getdnaSeqs()[0]); + + // V12346 mapped to A11111 + acf = protein.getCodonFrame(protein.getSequenceAt(1)).get(0); + assertEquals(1, acf.getdnaSeqs().length); + assertEquals(cdna.getSequenceAt(0).getDatasetSequence(), + acf.getdnaSeqs()[0]); + } + + /** + * Test the method that shows or hides sequence annotations by type(s) and + * selection group. + */ + @Test(groups = { "Functional" }) + public void testShowOrHideSequenceAnnotations() + { + SequenceI seq1 = new Sequence("Seq1", "AAA"); + SequenceI seq2 = new Sequence("Seq2", "BBB"); + SequenceI seq3 = new Sequence("Seq3", "CCC"); + Annotation[] anns = new Annotation[] { new Annotation(2f) }; + AlignmentAnnotation ann1 = new AlignmentAnnotation("Structure", "ann1", + anns); + ann1.setSequenceRef(seq1); + AlignmentAnnotation ann2 = new AlignmentAnnotation("Structure", "ann2", + anns); + ann2.setSequenceRef(seq2); + AlignmentAnnotation ann3 = new AlignmentAnnotation("Structure", "ann3", + anns); + AlignmentAnnotation ann4 = new AlignmentAnnotation("Temp", "ann4", anns); + ann4.setSequenceRef(seq1); + AlignmentAnnotation ann5 = new AlignmentAnnotation("Temp", "ann5", anns); + ann5.setSequenceRef(seq2); + AlignmentAnnotation ann6 = new AlignmentAnnotation("Temp", "ann6", anns); + AlignmentI al = new Alignment(new SequenceI[] { seq1, seq2, seq3 }); + al.addAnnotation(ann1); // Structure for Seq1 + al.addAnnotation(ann2); // Structure for Seq2 + al.addAnnotation(ann3); // Structure for no sequence + al.addAnnotation(ann4); // Temp for seq1 + al.addAnnotation(ann5); // Temp for seq2 + al.addAnnotation(ann6); // Temp for no sequence + List types = new ArrayList(); + List scope = new ArrayList(); + + /* + * Set all sequence related Structure to hidden (ann1, ann2) + */ + types.add("Structure"); + AlignmentUtils.showOrHideSequenceAnnotations(al, types, null, false, + false); + assertFalse(ann1.visible); + assertFalse(ann2.visible); + assertTrue(ann3.visible); // not sequence-related, not affected + assertTrue(ann4.visible); // not Structure, not affected + assertTrue(ann5.visible); // " + assertTrue(ann6.visible); // not sequence-related, not affected + + /* + * Set Temp in {seq1, seq3} to hidden + */ + types.clear(); + types.add("Temp"); + scope.add(seq1); + scope.add(seq3); + AlignmentUtils.showOrHideSequenceAnnotations(al, types, scope, false, + false); + assertFalse(ann1.visible); // unchanged + assertFalse(ann2.visible); // unchanged + assertTrue(ann3.visible); // not sequence-related, not affected + assertFalse(ann4.visible); // Temp for seq1 hidden + assertTrue(ann5.visible); // not in scope, not affected + assertTrue(ann6.visible); // not sequence-related, not affected + + /* + * Set Temp in all sequences to hidden + */ + types.clear(); + types.add("Temp"); + scope.add(seq1); + scope.add(seq3); + AlignmentUtils.showOrHideSequenceAnnotations(al, types, null, false, + false); + assertFalse(ann1.visible); // unchanged + assertFalse(ann2.visible); // unchanged + assertTrue(ann3.visible); // not sequence-related, not affected + assertFalse(ann4.visible); // Temp for seq1 hidden + assertFalse(ann5.visible); // Temp for seq2 hidden + assertTrue(ann6.visible); // not sequence-related, not affected + + /* + * Set all types in {seq1, seq3} to visible + */ + types.clear(); + scope.clear(); + scope.add(seq1); + scope.add(seq3); + AlignmentUtils.showOrHideSequenceAnnotations(al, types, scope, true, + true); + assertTrue(ann1.visible); // Structure for seq1 set visible + assertFalse(ann2.visible); // not in scope, unchanged + assertTrue(ann3.visible); // not sequence-related, not affected + assertTrue(ann4.visible); // Temp for seq1 set visible + assertFalse(ann5.visible); // not in scope, unchanged + assertTrue(ann6.visible); // not sequence-related, not affected + + /* + * Set all types in all scope to hidden + */ + AlignmentUtils.showOrHideSequenceAnnotations(al, types, null, true, + false); + assertFalse(ann1.visible); + assertFalse(ann2.visible); + assertTrue(ann3.visible); // not sequence-related, not affected + assertFalse(ann4.visible); + assertFalse(ann5.visible); + assertTrue(ann6.visible); // not sequence-related, not affected + } + + /** + * Tests for the method that checks if one sequence cross-references another + */ + @Test(groups = { "Functional" }) + public void testHasCrossRef() + { + assertFalse(AlignmentUtils.hasCrossRef(null, null)); + SequenceI seq1 = new Sequence("EMBL|A12345", "ABCDEF"); + assertFalse(AlignmentUtils.hasCrossRef(seq1, null)); + assertFalse(AlignmentUtils.hasCrossRef(null, seq1)); + SequenceI seq2 = new Sequence("UNIPROT|V20192", "ABCDEF"); + assertFalse(AlignmentUtils.hasCrossRef(seq1, seq2)); + + // different ref + seq1.addDBRef(new DBRefEntry("UNIPROT", "1", "v20193")); + assertFalse(AlignmentUtils.hasCrossRef(seq1, seq2)); + + // case-insensitive; version number is ignored + seq1.addDBRef(new DBRefEntry("UNIPROT", "1", "v20192")); + assertTrue(AlignmentUtils.hasCrossRef(seq1, seq2)); + + // right case! + seq1.addDBRef(new DBRefEntry("UNIPROT", "1", "V20192")); + assertTrue(AlignmentUtils.hasCrossRef(seq1, seq2)); + // test is one-way only + assertFalse(AlignmentUtils.hasCrossRef(seq2, seq1)); + } + + /** + * Tests for the method that checks if either sequence cross-references the + * other + */ + @Test(groups = { "Functional" }) + public void testHaveCrossRef() + { + assertFalse(AlignmentUtils.hasCrossRef(null, null)); + SequenceI seq1 = new Sequence("EMBL|A12345", "ABCDEF"); + assertFalse(AlignmentUtils.haveCrossRef(seq1, null)); + assertFalse(AlignmentUtils.haveCrossRef(null, seq1)); + SequenceI seq2 = new Sequence("UNIPROT|V20192", "ABCDEF"); + assertFalse(AlignmentUtils.haveCrossRef(seq1, seq2)); + + seq1.addDBRef(new DBRefEntry("UNIPROT", "1", "V20192")); + assertTrue(AlignmentUtils.haveCrossRef(seq1, seq2)); + // next is true for haveCrossRef, false for hasCrossRef + assertTrue(AlignmentUtils.haveCrossRef(seq2, seq1)); + + // now the other way round + seq1.setDBRefs(null); + seq2.addDBRef(new DBRefEntry("EMBL", "1", "A12345")); + assertTrue(AlignmentUtils.haveCrossRef(seq1, seq2)); + assertTrue(AlignmentUtils.haveCrossRef(seq2, seq1)); + + // now both ways + seq1.addDBRef(new DBRefEntry("UNIPROT", "1", "V20192")); + assertTrue(AlignmentUtils.haveCrossRef(seq1, seq2)); + assertTrue(AlignmentUtils.haveCrossRef(seq2, seq1)); + } + + /** + * Test the method that extracts the cds-only part of a dna alignment. + */ + @Test(groups = { "Functional" }) + public void testMakeCdsAlignment() + { + SequenceI dna1 = new Sequence("dna1", "aaaGGGcccTTTaaa"); + SequenceI dna2 = new Sequence("dna2", "GGGcccTTTaaaCCC"); + SequenceI pep1 = new Sequence("pep1", "GF"); + SequenceI pep2 = new Sequence("pep2", "GFP"); + 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); + + List mappings = new ArrayList(); + 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); + 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); + + AlignmentI cds = AlignmentUtils.makeCdsAlignment(new SequenceI[] { + dna1, dna2 }, mappings, dna); + assertEquals(2, cds.getSequences().size()); + assertEquals("---GGG---TTT---", cds.getSequenceAt(0) + .getSequenceAsString()); + assertEquals("GGG---TTT---CCC", 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 updated mappings + */ + assertEquals(2, mappings.size()); + + /* + * Mapping from pep1 to GGGTTT in first new exon sequence + */ + List pep1Mapping = MappingUtils + .findMappingsForSequence(pep1, mappings); + assertEquals(1, pep1Mapping.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()); + 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()); + assertEquals(4, m.getStart()); + assertEquals(6, m.getEnd()); + + /* + * Mapping from pep2 to GGGTTTCCC in second new exon sequence + */ + List pep2Mapping = MappingUtils + .findMappingsForSequence(pep2, mappings); + assertEquals(1, pep2Mapping.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()); + 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()); + 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()); + 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. + */ + @Test(groups = { "Functional" }) + public void testMakeCdsAlignment_multipleProteins() + { + SequenceI dna1 = new Sequence("dna1", "aaaGGGcccTTTaaa"); + SequenceI pep1 = new Sequence("pep1", "GF"); // GGGTTT + SequenceI pep2 = new Sequence("pep2", "KP"); // aaaccc + SequenceI pep3 = new Sequence("pep3", "KF"); // aaaTTT + dna1.createDatasetSequence(); + 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( + new DBRefEntry("EMBLCDS", "3", "A12346")); + pep3.getDatasetSequence().addDBRef( + new DBRefEntry("EMBLCDS", "4", "A12347")); + + /* + * 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); + + // 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); + + // 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); + + /* + * Create the Exon alignment; also replaces the dna-to-protein mappings with + * exon-to-protein and exon-to-dna mappings + */ + AlignmentI dna = new Alignment(new SequenceI[] { dna1 }); + dna.setDataset(null); + AlignmentI exal = AlignmentUtils.makeCdsAlignment( + new SequenceI[] { dna1 }, mappings, dna); + + /* + * Verify we have 3 cds sequences, mapped to pep1/2/3 respectively + */ + List cds = exal.getSequences(); + assertEquals(3, cds.size()); + + /* + * verify shared, extended alignment dataset + */ + assertSame(exal.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()); + + 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()); + + 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()); + + /* + * Verify there are mappings from each cds sequence to its protein product + * and also to its dna source + */ + Iterator newMappingsIterator = mappings.iterator(); + + // 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) + .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()); + } + + @Test(groups = { "Functional" }) + public void testIsMappable() + { + SequenceI dna1 = new Sequence("dna1", "cgCAGtgGT"); + SequenceI aa1 = new Sequence("aa1", "RSG"); + AlignmentI al1 = new Alignment(new SequenceI[] { dna1 }); + AlignmentI al2 = new Alignment(new SequenceI[] { aa1 }); + + assertFalse(AlignmentUtils.isMappable(null, null)); + assertFalse(AlignmentUtils.isMappable(al1, null)); + assertFalse(AlignmentUtils.isMappable(null, al1)); + assertFalse(AlignmentUtils.isMappable(al1, al1)); + assertFalse(AlignmentUtils.isMappable(al2, al2)); + + assertTrue(AlignmentUtils.isMappable(al1, al2)); + assertTrue(AlignmentUtils.isMappable(al2, al1)); + } + + /** + * Test creating a mapping when the sequences involved do not start at residue + * 1 + * + * @throws IOException + */ + @Test(groups = { "Functional" }) + public void testMapCdnaToProtein_forSubsequence() + throws IOException + { + SequenceI prot = new Sequence("UNIPROT|V12345", "E-I--Q", 10, 12); + prot.createDatasetSequence(); + + SequenceI dna = new Sequence("EMBL|A33333", "GAA--AT-C-CAG", 40, 48); + dna.createDatasetSequence(); + + MapList map = AlignmentUtils.mapCdnaToProtein(prot, dna); + assertEquals(10, map.getToLowest()); + assertEquals(12, map.getToHighest()); + assertEquals(40, map.getFromLowest()); + assertEquals(48, map.getFromHighest()); + } + + /** + * Test for the alignSequenceAs method where we have protein mapped to protein + */ + @Test(groups = { "Functional" }) + public void testAlignSequenceAs_mappedProteinProtein() + { + + SequenceI alignMe = new Sequence("Match", "MGAASEV"); + alignMe.createDatasetSequence(); + SequenceI alignFrom = new Sequence("Query", "LQTGYMGAASEVMFSPTRR"); + alignFrom.createDatasetSequence(); + + AlignedCodonFrame acf = new AlignedCodonFrame(); + // this is like a domain or motif match of part of a peptide sequence + MapList map = new MapList(new int[] { 6, 12 }, new int[] { 1, 7 }, 1, 1); + acf.addMap(alignFrom.getDatasetSequence(), + alignMe.getDatasetSequence(), map); + + AlignmentUtils.alignSequenceAs(alignMe, alignFrom, acf, "-", '-', true, + true); + assertEquals("-----MGAASEV-------", alignMe.getSequenceAsString()); + } + + /** + * Test for the alignSequenceAs method where there are trailing unmapped + * residues in the model sequence + */ + @Test(groups = { "Functional" }) + public void testAlignSequenceAs_withTrailingPeptide() + { + // map first 3 codons to KPF; G is a trailing unmapped residue + MapList map = new MapList(new int[] { 1, 9 }, new int[] { 1, 3 }, 3, 1); + + checkAlignSequenceAs("AAACCCTTT", "K-PFG", true, true, map, + "AAA---CCCTTT---"); + } + + /** + * Tests for transferring features between mapped sequences + */ + @Test(groups = { "Functional" }) + public void testTransferFeatures() + { + SequenceI dna = new Sequence("dna/20-34", "acgTAGcaaGCCcgt"); + SequenceI cds = new Sequence("cds/10-15", "TAGGCC"); + + // no overlap + dna.addSequenceFeature(new SequenceFeature("type1", "desc1", 1, 2, 1f, + null)); + // partial overlap - to [1, 1] + dna.addSequenceFeature(new SequenceFeature("type2", "desc2", 3, 4, 2f, + null)); + // exact overlap - to [1, 3] + dna.addSequenceFeature(new SequenceFeature("type3", "desc3", 4, 6, 3f, + null)); + // spanning overlap - to [2, 5] + dna.addSequenceFeature(new SequenceFeature("type4", "desc4", 5, 11, 4f, + null)); + // exactly overlaps whole mapped range [1, 6] + dna.addSequenceFeature(new SequenceFeature("type5", "desc5", 4, 12, 5f, + null)); + // no overlap (internal) + dna.addSequenceFeature(new SequenceFeature("type6", "desc6", 7, 9, 6f, + null)); + // no overlap (3' end) + dna.addSequenceFeature(new SequenceFeature("type7", "desc7", 13, 15, + 7f, null)); + // overlap (3' end) - to [6, 6] + dna.addSequenceFeature(new SequenceFeature("type8", "desc8", 12, 12, + 8f, null)); + // extended overlap - to [6, +] + dna.addSequenceFeature(new SequenceFeature("type9", "desc9", 12, 13, + 9f, null)); + + MapList map = new MapList(new int[] { 4, 6, 10, 12 }, + new int[] { 1, 6 }, 1, 1); + + /* + * transferFeatures() will build 'partial overlap' for regions + * that partially overlap 5' or 3' (start or end) of target sequence + */ + AlignmentUtils.transferFeatures(dna, cds, map, null); + SequenceFeature[] sfs = cds.getSequenceFeatures(); + assertEquals(6, sfs.length); + + SequenceFeature sf = sfs[0]; + assertEquals("type2", sf.getType()); + assertEquals("desc2", sf.getDescription()); + assertEquals(2f, sf.getScore()); + assertEquals(1, sf.getBegin()); + assertEquals(1, sf.getEnd()); + + sf = sfs[1]; + assertEquals("type3", sf.getType()); + assertEquals("desc3", sf.getDescription()); + assertEquals(3f, sf.getScore()); + assertEquals(1, sf.getBegin()); + assertEquals(3, sf.getEnd()); + + sf = sfs[2]; + assertEquals("type4", sf.getType()); + assertEquals(2, sf.getBegin()); + assertEquals(5, sf.getEnd()); + + sf = sfs[3]; + assertEquals("type5", sf.getType()); + assertEquals(1, sf.getBegin()); + assertEquals(6, sf.getEnd()); + + sf = sfs[4]; + assertEquals("type8", sf.getType()); + assertEquals(6, sf.getBegin()); + assertEquals(6, sf.getEnd()); + + sf = sfs[5]; + assertEquals("type9", sf.getType()); + assertEquals(6, sf.getBegin()); + assertEquals(6, sf.getEnd()); + } + + /** + * Tests for transferring features between mapped sequences + */ + @Test(groups = { "Functional" }) + public void testTransferFeatures_withOmit() + { + SequenceI dna = new Sequence("dna/20-34", "acgTAGcaaGCCcgt"); + SequenceI cds = new Sequence("cds/10-15", "TAGGCC"); + + MapList map = new MapList(new int[] { 4, 6, 10, 12 }, + new int[] { 1, 6 }, 1, 1); + + // [5, 11] maps to [2, 5] + dna.addSequenceFeature(new SequenceFeature("type4", "desc4", 5, 11, 4f, + null)); + // [4, 12] maps to [1, 6] + dna.addSequenceFeature(new SequenceFeature("type5", "desc5", 4, 12, 5f, + null)); + // [12, 12] maps to [6, 6] + dna.addSequenceFeature(new SequenceFeature("type8", "desc8", 12, 12, + 8f, null)); + + // desc4 and desc8 are the 'omit these' varargs + AlignmentUtils.transferFeatures(dna, cds, map, null, "type4", "type8"); + SequenceFeature[] sfs = cds.getSequenceFeatures(); + assertEquals(1, sfs.length); + + SequenceFeature sf = sfs[0]; + assertEquals("type5", sf.getType()); + assertEquals(1, sf.getBegin()); + assertEquals(6, sf.getEnd()); + } + + /** + * Tests for transferring features between mapped sequences + */ + @Test(groups = { "Functional" }) + public void testTransferFeatures_withSelect() + { + SequenceI dna = new Sequence("dna/20-34", "acgTAGcaaGCCcgt"); + SequenceI cds = new Sequence("cds/10-15", "TAGGCC"); + + MapList map = new MapList(new int[] { 4, 6, 10, 12 }, + new int[] { 1, 6 }, 1, 1); + + // [5, 11] maps to [2, 5] + dna.addSequenceFeature(new SequenceFeature("type4", "desc4", 5, 11, 4f, + null)); + // [4, 12] maps to [1, 6] + dna.addSequenceFeature(new SequenceFeature("type5", "desc5", 4, 12, 5f, + null)); + // [12, 12] maps to [6, 6] + dna.addSequenceFeature(new SequenceFeature("type8", "desc8", 12, 12, + 8f, null)); + + // "type5" is the 'select this type' argument + AlignmentUtils.transferFeatures(dna, cds, map, "type5"); + SequenceFeature[] sfs = cds.getSequenceFeatures(); + assertEquals(1, sfs.length); + + SequenceFeature sf = sfs[0]; + assertEquals("type5", sf.getType()); + assertEquals(1, sf.getBegin()); + assertEquals(6, sf.getEnd()); + } + + /** + * Test the method that extracts the cds-only part of a dna alignment, for the + * case where the cds should be aligned to match its nucleotide sequence. + */ + @Test(groups = { "Functional" }) + public void testMakeCdsAlignment_alternativeTranscripts() + { + SequenceI dna1 = new Sequence("dna1", "aaaGGGCC-----CTTTaaaGGG"); + // alternative transcript of same dna skips CCC codon + SequenceI dna2 = new Sequence("dna2", "aaaGGGCC-----cttTaaaGGG"); + // dna3 has no mapping (protein product) so should be ignored here + SequenceI dna3 = new Sequence("dna3", "aaaGGGCCCCCGGGcttTaaaGGG"); + SequenceI pep1 = new Sequence("pep1", "GPFG"); + SequenceI pep2 = new Sequence("pep2", "GPG"); + dna1.createDatasetSequence(); + dna2.createDatasetSequence(); + 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)); + + 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); + 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); + + AlignmentI dna = new Alignment(new SequenceI[] { dna1, dna2, dna3 }); + dna.setDataset(null); + AlignmentI cds = AlignmentUtils.makeCdsAlignment(new SequenceI[] { + dna1, dna2, dna3 }, mappings, dna); + List cdsSeqs = cds.getSequences(); + assertEquals(2, cdsSeqs.size()); + assertEquals("GGGCCCTTTGGG", cdsSeqs.get(0).getSequenceAsString()); + assertEquals("GGGCC---TGGG", cdsSeqs.get(1).getSequenceAsString()); + + /* + * 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 updated mappings + */ + assertEquals(2, mappings.size()); + + /* + * Mapping from pep1 to GGGTTT in first new CDS sequence + */ + List pep1Mapping = MappingUtils + .findMappingsForSequence(pep1, mappings); + assertEquals(1, pep1Mapping.size()); + /* + * maps GPFG to 1-3,4-6,7-9,10-12 + */ + SearchResults sr = MappingUtils.buildSearchResults(pep1, 1, mappings); + 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); + m = sr.getResults().get(0); + assertEquals(4, m.getStart()); + assertEquals(6, m.getEnd()); + sr = MappingUtils.buildSearchResults(pep1, 3, mappings); + m = sr.getResults().get(0); + assertEquals(7, m.getStart()); + assertEquals(9, m.getEnd()); + sr = MappingUtils.buildSearchResults(pep1, 4, mappings); + 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 + */ + List pep2Mapping = MappingUtils + .findMappingsForSequence(pep2, mappings); + assertEquals(1, pep2Mapping.size()); + sr = MappingUtils.buildSearchResults(pep2, 1, mappings); + 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); + m = sr.getResults().get(0); + assertEquals(4, m.getStart()); + assertEquals(6, m.getEnd()); + sr = MappingUtils.buildSearchResults(pep2, 3, mappings); + 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" }) + public void testAlignProteinAsDna_incompleteStartCodon() + { + // seq1: incomplete start codon (not mapped), then [3, 11] + SequenceI dna1 = new Sequence("Seq1", "ccAAA-TTT-GGG-"); + // seq2 codons are [4, 5], [8, 11] + SequenceI dna2 = new Sequence("Seq2", "ccaAA-ttT-GGG-"); + // seq3 incomplete start codon at 'tt' + SequenceI dna3 = new Sequence("Seq3", "ccaaa-ttt-GGG-"); + AlignmentI dna = new Alignment(new SequenceI[] { dna1, dna2, dna3 }); + dna.setDataset(null); + + // prot1 has 'X' for incomplete start codon (not mapped) + SequenceI prot1 = new Sequence("Seq1", "XKFG"); // X for incomplete start + SequenceI prot2 = new Sequence("Seq2", "NG"); + SequenceI prot3 = new Sequence("Seq3", "XG"); // X for incomplete start + AlignmentI protein = new Alignment(new SequenceI[] { prot1, prot2, + prot3 }); + protein.setDataset(null); + + // map dna1 [3, 11] to prot1 [2, 4] KFG + MapList map = new MapList(new int[] { 3, 11 }, new int[] { 2, 4 }, 3, 1); + AlignedCodonFrame acf = new AlignedCodonFrame(); + acf.addMap(dna1.getDatasetSequence(), prot1.getDatasetSequence(), map); + + // map dna2 [4, 5] [8, 11] to prot2 [1, 2] NG + map = new MapList(new int[] { 4, 5, 8, 11 }, new int[] { 1, 2 }, 3, 1); + acf.addMap(dna2.getDatasetSequence(), prot2.getDatasetSequence(), map); + + // map dna3 [9, 11] to prot3 [2, 2] G + map = new MapList(new int[] { 9, 11 }, new int[] { 2, 2 }, 3, 1); + acf.addMap(dna3.getDatasetSequence(), prot3.getDatasetSequence(), map); + + ArrayList acfs = new ArrayList(); + acfs.add(acf); + protein.setCodonFrames(acfs); + + /* + * verify X is included in the aligned proteins, and placed just + * before the first mapped residue + * CCT is between CCC and TTT + */ + AlignmentUtils.alignProteinAsDna(protein, dna); + assertEquals("XK-FG", prot1.getSequenceAsString()); + 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 testGetCdsRanges_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 testGetCdsRanges() + { + SequenceI dnaSeq = new Sequence("dna", "aaaGGGcccAAATTTttt"); + dnaSeq.createDatasetSequence(); + SequenceI ds = dnaSeq.getDatasetSequence(); + + // CDS for dna 3-6 + SequenceFeature sf = new SequenceFeature("CDS", "", 4, 6, 0f, null); + 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); + ds.addSequenceFeature(sf); + + List ranges = AlignmentUtils.findCdsPositions(dnaSeq); + 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 sf = new SequenceFeature("sequence_variant", "", 1, 1, + 0f, null); + sf.setValue("alleles", "T"); + dna.addSequenceFeature(sf); + + // two alleles codon 2, on bases 2 and 3 + sf = new SequenceFeature("sequence_variant", "", 5, 5, 0f, null); + sf.setValue("alleles", "T"); + dna.addSequenceFeature(sf); + sf = new SequenceFeature("sequence_variant", "", 6, 6, 0f, null); + sf.setValue("alleles", "G"); + dna.addSequenceFeature(sf); + + // two alleles codon 3, both on base 2 + sf = new SequenceFeature("sequence_variant", "", 8, 8, 0f, null); + sf.setValue("alleles", "C, G"); + dna.addSequenceFeature(sf); + + // no alleles on codon 4 + // alleles on codon 5 on all 3 bases + sf = new SequenceFeature("sequence_variant", "", 13, 13, 0f, null); + sf.setValue("alleles", "C, G"); // (C duplicates given base value) + dna.addSequenceFeature(sf); + sf = new SequenceFeature("sequence_variant", "", 14, 14, 0f, null); + sf.setValue("alleles", "g, a"); // should force to upper-case + dna.addSequenceFeature(sf); + sf = new SequenceFeature("sequence_variant", "", 15, 15, 0f, null); + sf.setValue("alleles", "A, T"); + dna.addSequenceFeature(sf); + + variantsMap = AlignmentUtils.buildDnaVariantsMap(dna, map); + assertEquals(4, variantsMap.size()); + assertTrue(Arrays.deepEquals(new String[][] { { "A", "T" }, { "T" }, + { "G" } }, variantsMap.get(1))); + assertTrue(Arrays.deepEquals(new String[][] { { "A" }, { "A", "T" }, + { "A", "G" } }, variantsMap.get(2))); + assertTrue(Arrays.deepEquals(new String[][] { { "T" }, + { "T", "C", "G" }, { "T" } }, variantsMap.get(3))); + // duplicated bases are not removed here, handled in computePeptideVariants + assertTrue(Arrays.deepEquals(new String[][] { { "C", "C", "G" }, + { "C", "G", "A" }, { "C", "A", "T" } }, variantsMap.get(5))); + } + + /** + * Tests for the method that computes all peptide variants given codon + * variants + */ + @Test(groups = "Functional") + public void testComputePeptideVariants() + { + String[][] codonVariants = new String[][] { { "A" }, { "G" }, { "T" } }; + + /* + * AGT codes for S - this is not included in the variants returned + */ + List variants = AlignmentUtils.computePeptideVariants(codonVariants, "S"); + assertEquals("[]", variants.toString()); + + // S is reported if it differs from the current value (A): + variants = AlignmentUtils.computePeptideVariants(codonVariants, "A"); + assertEquals("[S]", variants.toString()); + + /* + * synonymous variant is not reported + */ + codonVariants = new String[][] { { "A" }, { "G" }, { "C", "T" } }; + // AGC and AGT both code for S + variants = AlignmentUtils.computePeptideVariants(codonVariants, "s"); + assertEquals("[]", variants.toString()); + + /* + * equivalent variants are only reported once + */ + codonVariants = new String[][] { { "C" }, { "T" }, + { "A", "C", "G", "T" } }; + // CTA CTC CTG CTT all code for L + variants = AlignmentUtils.computePeptideVariants(codonVariants, "S"); + assertEquals("[L]", variants.toString()); + + /* + * vary codons 1 and 2; variant products are sorted and non-redundant + */ + codonVariants = new String[][] { { "a", "C" }, { "g", "T" }, { "A" } }; + // aga ata cga cta code for R, I, R, L + variants = AlignmentUtils.computePeptideVariants(codonVariants, "S"); + assertEquals("[I, L, R]", variants.toString()); + + /* + * vary codons 2 and 3 + */ + codonVariants = new String[][] { { "a" }, { "g", "T" }, { "A", "c" } }; + // aga agc ata atc code for R, S, I, I + variants = AlignmentUtils.computePeptideVariants(codonVariants, "S"); + assertEquals("[I, R]", variants.toString()); + + /* + * vary codons 1 and 3 + */ + codonVariants = new String[][] { { "a", "t" }, { "a" }, { "t", "g" } }; + // aat aag tat tag code for N, K, Y, STOP - STOP sorted to end + variants = AlignmentUtils.computePeptideVariants(codonVariants, "S"); + assertEquals("[K, N, Y, STOP]", variants.toString()); + + /* + * vary codons 1, 2 and 3 + */ + codonVariants = new String[][] { { "a", "t" }, { "G", "C" }, + { "t", "g" } }; + // agt agg act acg tgt tgg tct tcg code for S, R, T, T, C, W, S, S + variants = AlignmentUtils.computePeptideVariants(codonVariants, "S"); + assertEquals("[C, R, T, W]", variants.toString()); } }