package jalview.ext.ensembl; import static org.testng.AssertJUnit.assertEquals; import static org.testng.AssertJUnit.assertFalse; import static org.testng.AssertJUnit.assertNull; import static org.testng.AssertJUnit.assertTrue; import jalview.datamodel.SequenceDummy; import jalview.datamodel.SequenceFeature; import jalview.datamodel.SequenceI; import jalview.io.gff.SequenceOntologyFactory; import jalview.io.gff.SequenceOntologyLite; import jalview.util.MapList; import java.util.List; import org.testng.Assert; import org.testng.annotations.AfterClass; import org.testng.annotations.BeforeClass; import org.testng.annotations.Test; public class EnsemblCdnaTest { @BeforeClass(alwaysRun = true) public void setUp() { SequenceOntologyFactory.setInstance(new SequenceOntologyLite()); } @AfterClass(alwaysRun = true) public void tearDown() { SequenceOntologyFactory.setInstance(null); } /** * Test that the cdna part of genomic sequence is correctly identified by * 'exon' features (or subtypes) - reverse strand case. */ @Test(groups = "Functional") public void testGetGenomicRangesFromFeatures_reverseStrand() { EnsemblCdna testee = new EnsemblCdna(); SequenceI genomic = new SequenceDummy("chr7"); genomic.setStart(10000); genomic.setEnd(50000); String transcriptId = "ABC123"; // exon at (start+10000) length 501 SequenceFeature sf = new SequenceFeature("exon", "", 20000, 20500, 0f, null); sf.setValue("Parent", "transcript:" + transcriptId); sf.setStrand("-"); genomic.addSequenceFeature(sf); // exon (sub-type) at (start + exon_variant) length 101 sf = new SequenceFeature("coding_exon", "", 10500, 10600, 0f, null); sf.setValue("Parent", "transcript:" + transcriptId); sf.setStrand("-"); genomic.addSequenceFeature(sf); // exon belonging to a different transcript doesn't count sf = new SequenceFeature("exon", "", 11500, 12600, 0f, null); sf.setValue("Parent", "transcript:anotherOne"); genomic.addSequenceFeature(sf); // transcript feature doesn't count sf = new SequenceFeature("transcript", "", 10000, 50000, 0f, null); genomic.addSequenceFeature(sf); MapList ranges = testee.getGenomicRangesFromFeatures(genomic, transcriptId, 23); List fromRanges = ranges.getFromRanges(); assertEquals(2, fromRanges.size()); /* * from ranges should be sorted by start order descending * and hold mappings from reverse strand sense */ assertEquals(20500, fromRanges.get(0)[0]); assertEquals(20000, fromRanges.get(0)[1]); assertEquals(10600, fromRanges.get(1)[0]); assertEquals(10500, fromRanges.get(1)[1]); // to range should start from given start numbering List toRanges = ranges.getToRanges(); assertEquals(1, toRanges.size()); assertEquals(23, toRanges.get(0)[0]); assertEquals(624, toRanges.get(0)[1]); } /** * Test that the cdna part of genomic sequence is correctly identified by * 'exon' features (or subtypes) with the desired transcript as parent */ @Test(groups = "Functional") public void testGetGenomicRangesFromFeatures() { EnsemblCdna testee = new EnsemblCdna(); SequenceI genomic = new SequenceDummy("chr7"); genomic.setStart(10000); genomic.setEnd(50000); String transcriptId = "ABC123"; // exon at (start+10000) length 501 SequenceFeature sf = new SequenceFeature("exon", "", 20000, 20500, 0f, null); sf.setValue("Parent", "transcript:" + transcriptId); sf.setStrand("+"); genomic.addSequenceFeature(sf); // exon (sub-type) at (start + exon_variant) length 101 sf = new SequenceFeature("coding_exon", "", 10500, 10600, 0f, null); sf.setValue("Parent", "transcript:" + transcriptId); sf.setStrand("+"); genomic.addSequenceFeature(sf); // exon belonging to a different transcript doesn't count sf = new SequenceFeature("exon", "", 11500, 12600, 0f, null); sf.setValue("Parent", "transcript:anotherOne"); genomic.addSequenceFeature(sf); // transcript feature doesn't count sf = new SequenceFeature("transcript", "", 10000, 50000, 0f, null); sf.setStrand("-"); // weird but ignored genomic.addSequenceFeature(sf); MapList ranges = testee.getGenomicRangesFromFeatures(genomic, transcriptId, 23); List fromRanges = ranges.getFromRanges(); assertEquals(2, fromRanges.size()); // from ranges should be sorted by start order assertEquals(10500, fromRanges.get(0)[0]); assertEquals(10600, fromRanges.get(0)[1]); assertEquals(20000, fromRanges.get(1)[0]); assertEquals(20500, fromRanges.get(1)[1]); // to range should start from given start numbering List toRanges = ranges.getToRanges(); assertEquals(1, toRanges.size()); assertEquals(23, toRanges.get(0)[0]); assertEquals(624, toRanges.get(0)[1]); } /** * The method under test should give up and return null if both forward and * reverse strands are present in the features of interest */ @Test(groups = "Functional") public void testGetGenomicRangesFromFeatures_mixedStrand() { EnsemblCdna testee = new EnsemblCdna(); SequenceI genomic = new SequenceDummy("chr7"); genomic.setStart(10000); genomic.setEnd(50000); String transcriptId = "ABC123"; SequenceFeature sf = new SequenceFeature("exon", "", 20000, 20500, 0f, null); sf.setValue("Parent", "transcript:" + transcriptId); sf.setStrand("-"); genomic.addSequenceFeature(sf); sf = new SequenceFeature("coding_exon", "", 10500, 10600, 0f, null); sf.setValue("Parent", "transcript:" + transcriptId); sf.setStrand("+"); genomic.addSequenceFeature(sf); MapList ranges = testee.getGenomicRangesFromFeatures(genomic, transcriptId, 23); assertNull(ranges); } /** * Test the method that retains features except for 'transcript' (or * subtypes), or features with parent other than the given id */ @Test(groups = "Functional") public void testRetainFeature() { String accId = "ABC123"; EnsemblCdna testee = new EnsemblCdna(); SequenceFeature sf = new SequenceFeature("transcript", "", 20000, 20500, 0f, null); assertFalse(testee.retainFeature(sf, accId)); sf.setType("aberrant_processed_transcript"); assertFalse(testee.retainFeature(sf, accId)); sf.setType("NMD_transcript_variant"); assertFalse(testee.retainFeature(sf, accId)); // other feature with no parent is retained sf.setType("sequence_variant"); assertTrue(testee.retainFeature(sf, accId)); // other feature with desired parent is retained sf.setValue("Parent", "transcript:" + accId); assertTrue(testee.retainFeature(sf, accId)); // feature with wrong parent is not retained sf.setValue("Parent", "transcript:XYZ"); assertFalse(testee.retainFeature(sf, accId)); } /** * Test the method that picks out 'exon' (or subtype) features with the * accession id as parent */ @Test(groups = "Functional") public void testIdentifiesSequence() { String accId = "ABC123"; EnsemblCdna testee = new EnsemblCdna(); // exon with no parent not valid SequenceFeature sf = new SequenceFeature("exon", "", 1, 2, 0f, null); assertFalse(testee.identifiesSequence(sf, accId)); // exon with wrong parent not valid sf.setValue("Parent", "transcript:XYZ"); assertFalse(testee.identifiesSequence(sf, accId)); // exon with right parent is valid sf.setValue("Parent", "transcript:" + accId); assertTrue(testee.identifiesSequence(sf, accId)); // exon sub-type with right parent is valid sf.setType("coding_exon"); assertTrue(testee.identifiesSequence(sf, accId)); // transcript not valid: sf.setType("transcript"); assertFalse(testee.identifiesSequence(sf, accId)); // CDS not valid: sf.setType("CDS"); assertFalse(testee.identifiesSequence(sf, accId)); } @Test(groups = "Functional") public void testIsValidReference() throws Exception { EnsemblSequenceFetcher esq = new EnsemblCdna(); Assert.assertTrue(esq.isValidReference("CCDS5863.1")); Assert.assertTrue(esq.isValidReference("ENST00000288602")); Assert.assertTrue(esq.isValidReference("ENSG00000288602")); Assert.assertFalse(esq.isValidReference("ENSP00000288602")); Assert.assertFalse(esq.isValidReference("ENST0000288602")); // non-human species having a 3 character identifier included: Assert.assertTrue(esq.isValidReference("ENSMUSG00000099398")); } }