/*
- * Jalview - A Sequence Alignment Editor and Viewer (Version 2.8.2b1)
- * Copyright (C) 2014 The Jalview Authors
+ * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
+ * Copyright (C) $$Year-Rel$$ The Jalview Authors
*
* This file is part of Jalview.
*
*/
package jalview.analysis;
-import static org.junit.Assert.assertTrue;
-
-import org.junit.Test;
+import static org.testng.AssertJUnit.assertSame;
+import static org.testng.AssertJUnit.assertEquals;
+import static org.testng.AssertJUnit.assertFalse;
+import static org.testng.AssertJUnit.assertTrue;
+import static org.testng.AssertJUnit.assertNull;
+import org.testng.annotations.Test;
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.Collections;
+import java.util.HashSet;
+import java.util.Iterator;
+import java.util.LinkedHashSet;
+import java.util.List;
+import java.util.Map;
+import java.util.Set;
+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.SequenceI;
import jalview.io.AppletFormatAdapter;
+import jalview.io.FormatAdapter;
+import jalview.util.MapList;
+import jalview.util.MappingUtils;
public class AlignmentUtilsTests
{
- public static Sequence ts=new Sequence("short","ASDASDASDASDASDASDASDASDASDASDASDASDASD");
+ // @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
- public void testExpandFlanks()
+ 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);
al.addSequence(s1);
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
+ 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
+ public void testGetSequencesByName() throws IOException
+ {
+ final String data = ">Seq1Name\nKQYL\n" + ">Seq2Name\nRFPW\n"
+ + ">Seq1Name\nABCD\n";
+ AlignmentI al = loadAlignment(data, "FASTA");
+ Map<String, List<SequenceI>> 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
+ {
+ Alignment 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
+ public void testMapProteinToCdna_noXrefs() throws IOException
+ {
+ List<SequenceI> protseqs = new ArrayList<SequenceI>();
+ 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<SequenceI> dnaseqs = new ArrayList<SequenceI>();
+ 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.mapProteinToCdna(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
+ 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
+ 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
+ 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);
+
+
+ /*
+ * Expect alignment does nothing (aborts realignment). Change this test
+ * first if different behaviour wanted.
+ */
+ checkAlignSequenceAs("GGGAAACCCTTTGGG", "-A-L-P-", false,
+ false, map, "GGGAAACCCTTTGGG");
+ }
+
+ /**
+ * Helper method that performs and verifies the method under test.
+ *
+ * @param dnaSeq
+ * @param proteinSeq
+ * @param preserveMappedGaps
+ * @param preserveUnmappedGaps
+ * @param map
+ * @param expected
+ */
+ protected void checkAlignSequenceAs(final String dnaSeq,
+ final String proteinSeq, final boolean preserveMappedGaps,
+ final boolean preserveUnmappedGaps, MapList map,
+ final String expected)
+ {
+ SequenceI dna = new Sequence("Seq1", dnaSeq);
+ dna.createDatasetSequence();
+ SequenceI protein = new Sequence("Seq1", proteinSeq);
+ protein.createDatasetSequence();
+ AlignedCodonFrame acf = new AlignedCodonFrame();
+ acf.addMap(dna.getDatasetSequence(), protein.getDatasetSequence(), map);
+
+ AlignmentUtils.alignSequenceAs(dna, protein, acf, "---", '-',
+ preserveMappedGaps, preserveUnmappedGaps);
+ assertEquals(expected, dna.getSequenceAsString());
+ }
+
+ /**
+ * Test for the alignSequenceAs method where we preserve gaps in introns only.
+ */
+ @Test
+ 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
+ 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
+ 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");
+ AlignmentI protein = new Alignment(new SequenceI[]
+ { prot1, prot2, prot3 });
+ 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);
+ protein.setCodonFrames(Collections.singleton(acf));
+
+ /*
+ * 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());
+ }
+
+ /**
+ * Test the method that tests whether a CDNA sequence translates to a protein
+ * sequence
+ */
+ @Test
+ public void testTranslatesAs()
+ {
+ assertTrue(AlignmentUtils.translatesAs("tttcccaaaggg".toCharArray(), 0,
+ "FPKG".toCharArray()));
+ // with start codon
+ assertTrue(AlignmentUtils.translatesAs("atgtttcccaaaggg".toCharArray(),
+ 3, "FPKG".toCharArray()));
+ // with stop codon1
+ assertTrue(AlignmentUtils.translatesAs("tttcccaaagggtaa".toCharArray(),
+ 0, "FPKG".toCharArray()));
+ // with stop codon2
+ assertTrue(AlignmentUtils.translatesAs("tttcccaaagggtag".toCharArray(),
+ 0, "FPKG".toCharArray()));
+ // with stop codon3
+ assertTrue(AlignmentUtils.translatesAs("tttcccaaagggtga".toCharArray(),
+ 0, "FPKG".toCharArray()));
+ // with start and stop codon1
+ assertTrue(AlignmentUtils.translatesAs(
+ "atgtttcccaaaggtaa".toCharArray(), 3, "FPKG".toCharArray()));
+ // with start and stop codon2
+ assertTrue(AlignmentUtils.translatesAs(
+ "atgtttcccaaaggtag".toCharArray(), 3, "FPKG".toCharArray()));
+ // with start and stop codon3
+ assertTrue(AlignmentUtils.translatesAs(
+ "atgtttcccaaaggtga".toCharArray(), 3, "FPKG".toCharArray()));
+
+ // wrong protein
+ assertFalse(AlignmentUtils.translatesAs("tttcccaaaggg".toCharArray(),
+ 0,
+ "FPMG".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
+ public void testMapProteinToCdna_withStartAndStopCodons()
+ throws IOException
+ {
+ List<SequenceI> protseqs = new ArrayList<SequenceI>();
+ 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<SequenceI> dnaseqs = new ArrayList<SequenceI>();
+ // 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.mapProteinToCdna(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
+ public void testMapProteinToCdna_withXrefs() throws IOException
+ {
+ List<SequenceI> protseqs = new ArrayList<SequenceI>();
+ 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<SequenceI> dnaseqs = new ArrayList<SequenceI>();
+ 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.mapProteinToCdna(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
+ public void testMapProteinToCdna_prioritiseXrefs() throws IOException
+ {
+ List<SequenceI> protseqs = new ArrayList<SequenceI>();
+ 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<SequenceI> dnaseqs = new ArrayList<SequenceI>();
+ 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.mapProteinToCdna(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
+ 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<String> types = new ArrayList<String>();
+ List<SequenceI> scope = new ArrayList<SequenceI>();
+
+ /*
+ * 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
+ 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
+ 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.setDBRef(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 exon-only part of a dna alignment.
+ */
+ @Test
+ public void testMakeExonAlignment()
+ {
+ 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();
+
+ Set<AlignedCodonFrame> mappings = new HashSet<AlignedCodonFrame>();
+ 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 exons = AlignmentUtils.makeExonAlignment(new SequenceI[]
+ { dna1, dna2 }, mappings);
+ assertEquals(2, exons.getSequences().size());
+ assertEquals("GGGTTT", exons.getSequenceAt(0).getSequenceAsString());
+ assertEquals("GGGTTTCCC", exons.getSequenceAt(1).getSequenceAsString());
+
+ /*
+ * Verify updated mappings
+ */
+ assertEquals(2, mappings.size());
+
+ /*
+ * Mapping from pep1 to GGGTTT in first new exon sequence
+ */
+ List<AlignedCodonFrame> 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);
+ assertEquals(exons.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);
+ assertEquals(exons.getSequenceAt(0).getDatasetSequence(),
+ m.getSequence());
+ assertEquals(4, m.getStart());
+ assertEquals(6, m.getEnd());
+
+ /*
+ * Mapping from pep2 to GGGTTTCCC in second new exon sequence
+ */
+ List<AlignedCodonFrame> 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);
+ assertEquals(exons.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);
+ assertEquals(exons.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);
+ assertEquals(exons.getSequenceAt(1).getDatasetSequence(),
+ m.getSequence());
+ assertEquals(7, m.getStart());
+ assertEquals(9, m.getEnd());
+ }
+
+ /**
+ * Test the method that makes an exon-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
+ public void testMakeExonSequences()
+ {
+ 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<AlignedCodonFrame> mappings = new HashSet<AlignedCodonFrame>();
+ 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<SequenceI> exons = AlignmentUtils.makeExonSequences(dna1, acf,
+ newMapping);
+ assertEquals(1, exons.size());
+ SequenceI exon = exons.get(0);
+
+ assertEquals("GGGTTT", exon.getSequenceAsString());
+ assertEquals("dna1|A12345", exon.getName());
+ assertEquals(1, exon.getDBRef().length);
+ DBRefEntry cdsRef = exon.getDBRef()[0];
+ assertEquals("EMBLCDS", cdsRef.getSource());
+ assertEquals("2", cdsRef.getVersion());
+ assertEquals("A12345", cdsRef.getAccessionId());
+ }
+
+ /**
+ * Test the method that makes an exon-only alignment from a DNA sequence and
+ * its product mappings, for the case where there are multiple exon mappings
+ * to different protein products.
+ */
+ @Test
+ public void testMakeExonAlignment_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();
+ 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. Using LinkedHashset is a
+ * convenience so results are in the input order. There is no assertion that
+ * the generated exon sequences are in any particular order.
+ */
+ Set<AlignedCodonFrame> mappings = new LinkedHashSet<AlignedCodonFrame>();
+ // 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 exal = AlignmentUtils.makeExonAlignment(new SequenceI[]
+ { dna1 }, mappings);
+
+ /*
+ * Verify we have 3 exon sequences, mapped to pep1/2/3 respectively
+ */
+ List<SequenceI> exons = exal.getSequences();
+ assertEquals(3, exons.size());
+
+ SequenceI exon = exons.get(0);
+ assertEquals("GGGTTT", exon.getSequenceAsString());
+ assertEquals("dna1|A12345", exon.getName());
+ assertEquals(1, exon.getDBRef().length);
+ DBRefEntry cdsRef = exon.getDBRef()[0];
+ assertEquals("EMBLCDS", cdsRef.getSource());
+ assertEquals("2", cdsRef.getVersion());
+ assertEquals("A12345", cdsRef.getAccessionId());
+
+ exon = exons.get(1);
+ assertEquals("aaaccc", exon.getSequenceAsString());
+ assertEquals("dna1|A12346", exon.getName());
+ assertEquals(1, exon.getDBRef().length);
+ cdsRef = exon.getDBRef()[0];
+ assertEquals("EMBLCDS", cdsRef.getSource());
+ assertEquals("3", cdsRef.getVersion());
+ assertEquals("A12346", cdsRef.getAccessionId());
+
+ exon = exons.get(2);
+ assertEquals("aaaTTT", exon.getSequenceAsString());
+ assertEquals("dna1|A12347", exon.getName());
+ assertEquals(1, exon.getDBRef().length);
+ cdsRef = exon.getDBRef()[0];
+ assertEquals("EMBLCDS", cdsRef.getSource());
+ assertEquals("4", cdsRef.getVersion());
+ assertEquals("A12347", cdsRef.getAccessionId());
+
+ /*
+ * Verify there are mappings from each exon sequence to its protein product
+ * and also to its dna source
+ */
+ Iterator<AlignedCodonFrame> newMappingsIterator = mappings.iterator();
+
+ // mappings for dna1 - exon1 - pep1
+ AlignedCodonFrame exonMapping = newMappingsIterator.next();
+ List<Mapping> dnaMappings = exonMapping.getMappingsForSequence(dna1);
+ assertEquals(1, dnaMappings.size());
+ assertSame(exons.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<Mapping> peptideMappings = exonMapping
+ .getMappingsForSequence(pep1);
+ assertEquals(1, peptideMappings.size());
+ assertSame(pep1.getDatasetSequence(), peptideMappings.get(0).getTo());
+
+ // mappings for dna1 - exon2 - pep2
+ exonMapping = newMappingsIterator.next();
+ dnaMappings = exonMapping.getMappingsForSequence(dna1);
+ assertEquals(1, dnaMappings.size());
+ assertSame(exons.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 = exonMapping.getMappingsForSequence(pep2);
+ assertEquals(1, peptideMappings.size());
+ assertSame(pep2.getDatasetSequence(), peptideMappings.get(0).getTo());
+
+ // mappings for dna1 - exon3 - pep3
+ exonMapping = newMappingsIterator.next();
+ dnaMappings = exonMapping.getMappingsForSequence(dna1);
+ assertEquals(1, dnaMappings.size());
+ assertSame(exons.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 = exonMapping.getMappingsForSequence(pep3);
+ assertEquals(1, peptideMappings.size());
+ assertSame(pep3.getDatasetSequence(), peptideMappings.get(0).getTo());
}
}
git://source.jalview.org / jalview.git / blobdiff