test/jalview/ws/dbsources/UniprotTest.java

   1 /*
   2  * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
   3  * Copyright (C) $$Year-Rel$$ The Jalview Authors
   4  *
   5  * This file is part of Jalview.
   6  *
   7  * Jalview is free software: you can redistribute it and/or
   8  * modify it under the terms of the GNU General Public License
   9  * as published by the Free Software Foundation, either version 3
  10  * of the License, or (at your option) any later version.
  11  *
  12  * Jalview is distributed in the hope that it will be useful, but
  13  * WITHOUT ANY WARRANTY; without even the implied warranty
  14  * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
  15  * PURPOSE.  See the GNU General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU General Public License
  18  * along with Jalview.  If not, see <http://www.gnu.org/licenses/>.
  19  * The Jalview Authors are detailed in the 'AUTHORS' file.
  20  */
  21 package jalview.ws.dbsources;
  22
  23 import static org.testng.AssertJUnit.assertEquals;
  24 import static org.testng.AssertJUnit.assertNull;
  25
  26 import jalview.datamodel.PDBEntry;
  27 import jalview.datamodel.SequenceFeature;
  28 import jalview.datamodel.UniprotEntry;
  29
  30 import java.io.Reader;
  31 import java.io.StringReader;
  32 import java.util.Vector;
  33
  34 import org.testng.annotations.Test;
  35
  36 public class UniprotTest
  37 {
  38   // adapted from http://www.uniprot.org/uniprot/A9CKP4.xml
  39   private static final String UNIPROT_XML = "<?xml version='1.0' encoding='UTF-8'?>"
  40           + "<uniprot>"
  41           + "<entry dataset=\"TrEMBL\" created=\"2008-01-15\" modified=\"2015-03-04\" version=\"38\">"
  42           + "<accession>A9CKP4</accession>"
  43           + "<accession>A9CKP5</accession>"
  44           + "<name>A9CKP4_AGRT5</name>"
  45           + "<name>A9CKP4_AGRT6</name>"
  46           + "<protein><recommendedName><fullName>Mitogen-activated protein kinase 13</fullName><fullName>Henry</fullName></recommendedName></protein>"
  47           + "<dbReference type=\"PDB\" id=\"2FSQ\"><property type=\"method\" value=\"X-ray\"/><property type=\"resolution\" value=\"1.40\"/></dbReference>"
  48           + "<dbReference type=\"PDBsum\" id=\"2FSR\"/>"
  49           + "<feature type=\"signal peptide\" evidence=\"7\"><location><begin position=\"1\"/><end position=\"18\"/></location></feature>"
  50           + "<feature type=\"propeptide\" description=\"Activation peptide\" id=\"PRO_0000027399\" evidence=\"9 16 17 18\"><location><begin position=\"19\"/><end position=\"20\"/></location></feature>"
  51           + "<feature type=\"chain\" description=\"Granzyme B\" id=\"PRO_0000027400\"><location><begin position=\"21\"/><end position=\"247\"/></location></feature>"
  52           + "<sequence length=\"10\" mass=\"27410\" checksum=\"8CB760AACF88FE6C\" modified=\"2008-01-15\" version=\"1\">MHAPL VSKDL</sequence></entry>"
  53           + "</uniprot>";
  54
  55   /**
  56    * Test the method that unmarshals XML to a Uniprot model
  57    */
  58   @Test(groups = { "Functional" })
  59   public void testGetUniprotEntries()
  60   {
  61     Uniprot u = new Uniprot();
  62     Reader reader = new StringReader(UNIPROT_XML);
  63     Vector<UniprotEntry> entries = u.getUniprotEntries(reader);
  64     assertEquals(1, entries.size());
  65     UniprotEntry entry = entries.get(0);
  66     assertEquals(2, entry.getName().size());
  67     assertEquals("A9CKP4_AGRT5", entry.getName().get(0));
  68     assertEquals("A9CKP4_AGRT6", entry.getName().get(1));
  69     assertEquals(2, entry.getAccession().size());
  70     assertEquals("A9CKP4", entry.getAccession().get(0));
  71     assertEquals("A9CKP5", entry.getAccession().get(1));
  72
  73     /*
  74      * UniprotSequence drops any space characters
  75      */
  76     assertEquals("MHAPLVSKDL", entry.getUniprotSequence().getContent());
  77
  78     assertEquals(2, entry.getProtein().getName().size());
  79     assertEquals("Mitogen-activated protein kinase 13", entry.getProtein()
  80             .getName().get(0));
  81     assertEquals("Henry", entry.getProtein().getName().get(1));
  82
  83     /*
  84      * Check sequence features
  85      */
  86     Vector<SequenceFeature> features = entry.getFeature();
  87     assertEquals(3, features.size());
  88     SequenceFeature sf = features.get(0);
  89     assertEquals("signal peptide", sf.getType());
  90     assertNull(sf.getDescription());
  91     assertNull(sf.getStatus());
  92     assertEquals(1, sf.getPosition());
  93     assertEquals(1, sf.getBegin());
  94     assertEquals(18, sf.getEnd());
  95     sf = features.get(1);
  96     assertEquals("propeptide", sf.getType());
  97     assertEquals("Activation peptide", sf.getDescription());
  98     assertEquals(19, sf.getPosition());
  99     assertEquals(19, sf.getBegin());
 100     assertEquals(20, sf.getEnd());
 101     sf = features.get(2);
 102     assertEquals("chain", sf.getType());
 103     assertEquals("Granzyme B", sf.getDescription());
 104     assertEquals(21, sf.getPosition());
 105     assertEquals(21, sf.getBegin());
 106     assertEquals(247, sf.getEnd());
 107
 108     /*
 109      * Check cross-references
 110      */
 111     Vector<PDBEntry> xrefs = entry.getDbReference();
 112     assertEquals(2, xrefs.size());
 113
 114     PDBEntry xref = xrefs.get(0);
 115     assertEquals("2FSQ", xref.getId());
 116     assertEquals("PDB", xref.getType());
 117     assertEquals(2, xref.getProperty().size());
 118     assertEquals("X-ray", xref.getProperty().get("method"));
 119     assertEquals("1.40", xref.getProperty().get("resolution"));
 120
 121     xref = xrefs.get(1);
 122     assertEquals("2FSR", xref.getId());
 123     assertEquals("PDBsum", xref.getType());
 124     assertNull(xref.getProperty());
 125   }
 126
 127   /**
 128    * Test the method that formats the sequence id
 129    */
 130   @Test(groups = { "Functional" })
 131   public void testGetUniprotEntryId()
 132   {
 133     UniprotEntry entry = new Uniprot().getUniprotEntries(
 134             new StringReader(UNIPROT_XML)).get(0);
 135
 136     /*
 137      * name formatted as source | accession ids | names
 138      * source database converted to Jalview canonical name
 139      */
 140     String expectedName = "UNIPROT|A9CKP4|A9CKP5|A9CKP4_AGRT5|A9CKP4_AGRT6";
 141     assertEquals(expectedName, Uniprot.getUniprotEntryId(entry));
 142   }
 143
 144   /**
 145    * Test the method that formats the sequence description
 146    */
 147   @Test(groups = { "Functional" })
 148   public void testGetUniprotEntryDescription()
 149   {
 150     UniprotEntry entry = new Uniprot().getUniprotEntries(
 151             new StringReader(UNIPROT_XML)).get(0);
 152
 153     /*
 154      * recommended names concatenated with space separator
 155      */
 156     String expectedDescription = "Mitogen-activated protein kinase 13 Henry";
 157     assertEquals(expectedDescription,
 158             Uniprot.getUniprotEntryDescription(entry));
 159   }
 160 }