import static jalview.io.gff.SequenceOntologyI.SEQUENCE_VARIANT;
import static org.testng.Assert.assertEquals;
import static org.testng.Assert.assertNull;
-import static org.testng.Assert.assertSame;
import static org.testng.Assert.assertTrue;
import jalview.bin.Cache;
+ ">transcript4/1-18\n-----TGG-GGACGAGAGTGTGA-A\n";
private static final String[] VCF = { "##fileformat=VCFv4.2",
- // fields other than AF are ignored when parsing as they have no INFO definition
+ // note fields with no INFO definition are ignored when parsing
"##INFO=<ID=AF,Number=A,Type=Float,Description=\"Allele Frequency, for each ALT allele, in the same order as listed\">",
"##INFO=<ID=AC_Female,Number=A,Type=Integer,Description=\"Allele count in Female genotypes\"",
"##INFO=<ID=AF_AFR,Number=A,Type=Float,Description=\"Allele Frequency among African/African American genotypes\"",
+ "##INFO=<ID=CLNSIG,Number=.,Type=String,Description=\"Clinical significance for this single variant\"",
"##reference=Homo_sapiens/GRCh38",
"#CHROM\tPOS\tID\tREF\tALT\tQUAL\tFILTER\tINFO",
// A/T,C variants in position 2 of gene sequence (precedes transcript)
// should create 2 variant features with respective AF values
// malformed values for AC_Female and AF_AFR should be ignored
- "17\t45051611\trs384765\tA\tT,C\t1666.64\tRF;XYZ\tAC=15;AF=5.0e-03,4.0e-03;AC_Female=12,3d;AF_AFR=low,2.3e-4",
+ "17\t45051611\trs384765\tA\tT,C\t1666.64\tRF;XYZ\tAC=15;AF=5.0e-03,4.0e-03;AC_Female=12,3d;AF_AFR=low,2.3e-4;CLNSIG=benign,probably_benign",
// SNP G/C in position 4 of gene sequence, position 2 of transcript
// insertion G/GA is transferred to nucleotide but not to peptide
"17\t45051613\t.\tG\tGA,C\t1666.65\t.\tAC=15;AF=3.0e-03,2.0e-03",
assertEquals(sf.getValue("ID"), "rs384765");
assertEquals(sf.getValue("QUAL"), "1666.64");
assertEquals(sf.getValue("FILTER"), "RF;XYZ");
+ /*
+ * if INFO declares Number=1, all values are attached to each allele
+ */
+ assertEquals(sf.getValue("CLNSIG"), "benign,probably_benign");
// malformed integer for AC_Female is ignored (JAL-3375)
assertNull(sf.getValue("AC_Female"));
// malformed float for AF_AFR is ignored (JAL-3375)
assertNull(sf.getValue("AC_AFR"));
assertEquals(sf.getValue(Gff3Helper.ALLELES), "A,T");
+ assertEquals(sf.getValue("CLNSIG"), "benign,probably_benign");
sf = geneFeatures.get(2);
assertEquals(sf.getFeatureGroup(), "VCF");
assertEquals(sf.getValue("alleles"), "C,T");
map = (Map) sf.getValue("CSQ");
assertEquals(map.size(), 9);
- assertEquals(map.get("PolyPhen"), "Bad++"); // %3B%3B decoded
+ assertEquals(map.get("PolyPhen"), "Bad;;"); // %3B%3B decoded
sf = geneFeatures.get(2);
assertEquals(sf.getBegin(), 9);
assertEquals(sf.getEnd(), 15);
assertEquals(sf.getDescription(), "T,C");
}
-
- @Test(groups = "Functional")
- public void testDecodeSpecialCharacters() throws IOException
- {
- String encoded = "hello world";
- String decoded = VCFLoader.decodeSpecialCharacters(encoded);
- assertSame(encoded, decoded); // no change needed
-
- encoded = "ab%3Acd%3Bef%3Dgh%25ij%2Ckl%3A";
- decoded = VCFLoader.decodeSpecialCharacters(encoded);
- assertEquals(decoded, "ab:cd;ef=gh%ij,kl:");
- }
}
\ No newline at end of file