incorporate the cvtermpath test obo into tests

tests/tripal_chado/api/OBOImporterTest.php

@@ -1,50 +0,0 @@
-<?php
-namespace Tests\tripal_chado\api;
-
-use StatonLab\TripalTestSuite\DBTransaction;
-use StatonLab\TripalTestSuite\TripalTestCase;
-
-module_load_include('inc', 'tripal_chado', 'includes/TripalImporter/OBOImporter.inc');
-
-
-class OBOImporterTest extends TripalTestCase {
-  
-  use DBTransaction;
-    
-  /**
-   * Tests that the OBO loader can import from a remote OBO.  
-   *
-   * @group api
-   * @group chado
-   * @group obo
-   * 
-   */
-  public function testRemoteRemote() {
-    
-    // Make sure the sequence ontology OBO is there.
-    chado_insert_obo('sequence', 'http://purl.obolibrary.org/obo/so.obo');
-        
-    // The loader is an instance of TripalImporter which
-    // requires a job. So let's create one.
-    $so = new TripalJob;
-    $so->create([
-      'job_name' => 'OBO test',
-      'modulename' => 'tripal_chado',
-      'callback' => 'NA',
-      'arguments' => [
-        'obo_id' => chado_get_obo('sequence'),
-      ],
-      'uid' => 1,
-      'priority' => 10,
-      'includes' => []]);
-      
-      return [
-        $so,
-      ];
-    
-    $loader = new \OBOImporter($job);
-    $loader->run();
-    
-    $check_sql = "SELECT count(*) FROM " 
-  }
-}

tests/tripal_chado/example_files/go_custom_mini.obo

@@ -0,0 +1,190 @@
+format-version: 1.2
+data-version: releases/2018-08-28
+subsetdef: gocheck_do_not_annotate "Term not to be used for direct annotation"
+subsetdef: gocheck_do_not_manually_annotate "Term not to be used for direct manual annotation"
+subsetdef: goslim_agr "AGR slim"
+subsetdef: goslim_aspergillus "Aspergillus GO slim"
+subsetdef: goslim_candida "Candida GO slim"
+subsetdef: goslim_chembl "ChEMBL protein targets summary"
+subsetdef: goslim_generic "Generic GO slim"
+subsetdef: goslim_metagenomics "Metagenomics GO slim"
+subsetdef: goslim_mouse "Mouse GO slim"
+subsetdef: goslim_pir "PIR GO slim"
+subsetdef: goslim_plant "Plant GO slim"
+subsetdef: goslim_pombe "Fission yeast GO slim"
+subsetdef: goslim_synapse "synapse GO slim"
+subsetdef: goslim_yeast "Yeast GO slim"
+synonymtypedef: syngo_official_label "label approved by the SynGO project"
+synonymtypedef: systematic_synonym "Systematic synonym" EXACT
+default-namespace: gene_ontology
+remark: cvs version: $Revision: 38972$
+remark: Includes Ontology(OntologyID(Anonymous-39)) [Axioms: 277 Logical Axioms: 277]
+remark: Includes Ontology(OntologyID(OntologyIRI(<http://purl.obolibrary.org/obo/go/never_in_taxon.owl>))) [Axioms: 18 Logical Axioms: 0]
+ontology: go
+property_value: http://purl.org/dc/elements/1.1/license http://creativecommons.org/licenses/by/4.0/
+
+
+[Term]
+id: GO:0000156
+name: phosphorelay response regulator activity
+namespace: molecular_function
+def: "Responds to a phosphorelay sensor to initiate a change in cell state or activity. The activity of the response regulator is regulated by transfer of a phosphate from a histidine residue in the sensor, to an aspartate residue in the response regulator. Many but not all response regulators act as transcriptional regulators to elicit a response." [GOC:bf, PMID:10966457, PMID:11842140]
+comment: Added is_a molecular transducer activity see https://github.com/geneontology/go-ontology/issues/15611
+synonym: "two-component response regulator activity" NARROW []
+is_a: GO:0060089 ! molecular transducer activity
+relationship: part_of GO:0000160 ! phosphorelay signal transduction system
+
+
+[Term]
+id: GO:0004679
+name: AMP-activated protein kinase activity
+namespace: molecular_function
+def: "Catalysis of the reaction: ATP + a protein = ADP + a phosphoprotein. This reaction requires the presence of AMP." [GOC:mah]
+synonym: "5'-AMP-activated protein kinase activity" EXACT []
+synonym: "AMPK activity" RELATED [GOC:mah]
+synonym: "SNF1A/AMP-activated protein kinase activity" NARROW []
+xref: EC:2.7.11
+xref: Reactome:REACT_100101 "Phosphorylated AMPK phosphorylates TSC2, Danio rerio"
+xref: Reactome:REACT_100513 "Phosphorylated AMPK phosphorylates TSC2, Xenopus tropicalis"
+xref: Reactome:REACT_101641 "pAMPK inactivates ACC2 inhibiting malonyl-CoA synthesis, Arabidopsis thaliana"
+xref: Reactome:REACT_102231 "Phosphorylated AMPK phosphorylates TSC2, Bos taurus"
+xref: Reactome:REACT_104140 "pAMPK inactivates ACC2 inhibiting malonyl-CoA synthesis, Gallus gallus"
+xref: Reactome:REACT_105858 "pAMPK inactivates ACC2 inhibiting malonyl-CoA synthesis, Saccharomyces cerevisiae"
+xref: Reactome:REACT_107230 "Phosphorylation of ChREBP at Serine 568 by AMPK, Mus musculus"
+xref: Reactome:REACT_107271 "pAMPK inactivates ACC2 inhibiting malonyl-CoA synthesis, Mus musculus"
+xref: Reactome:REACT_107536 "Phosphorylated AMPK phosphorylates TSC2, Rattus norvegicus"
+xref: Reactome:REACT_11110 "pAMPK inactivates ACC2 inhibiting malonyl-CoA synthesis, Homo sapiens"
+xref: Reactome:REACT_113551 "Phosphorylation of ChREBP at Serine 568 by AMPK, Danio rerio"
+xref: Reactome:REACT_114056 "pAMPK inactivates ACC2 inhibiting malonyl-CoA synthesis, Danio rerio"
+xref: Reactome:REACT_115260 "pAMPK inactivates ACC2 inhibiting malonyl-CoA synthesis, Taeniopygia guttata"
+xref: Reactome:REACT_115507 "Phosphorylation of ChREBP at Serine 568 by AMPK, Xenopus tropicalis"
+xref: Reactome:REACT_1706 "Phosphorylation of rChREBP(Ser 568) by rAMPK, Rattus norvegicus"
+xref: Reactome:REACT_21348 "Phosphorylated AMPK phosphorylates TSC2, Homo sapiens"
+xref: Reactome:REACT_21413 "AMPK phosphorylates Raptor, Homo sapiens"
+xref: Reactome:REACT_28253 "pAMPK inactivates ACC2 inhibiting malonyl-CoA synthesis, Oryza sativa"
+xref: Reactome:REACT_31105 "Phosphorylation of ChREBP at Serine 568 by AMPK, Bos taurus"
+xref: Reactome:REACT_31214 "pAMPK inactivates ACC2 inhibiting malonyl-CoA synthesis, Canis familiaris"
+xref: Reactome:REACT_32411 "pAMPK inactivates ACC2 inhibiting malonyl-CoA synthesis, Drosophila melanogaster"
+xref: Reactome:REACT_32709 "pAMPK inactivates ACC2 inhibiting malonyl-CoA synthesis, Dictyostelium discoideum"
+xref: Reactome:REACT_33102 "Phosphorylation of ChREBP at Serine 568 by AMPK, Sus scrofa"
+xref: Reactome:REACT_349 "Phosphorylation of ChREBP at Serine 568 by AMPK, Homo sapiens"
+xref: Reactome:REACT_82669 "pAMPK inactivates ACC2 inhibiting malonyl-CoA synthesis, Rattus norvegicus"
+xref: Reactome:REACT_85819 "Phosphorylation of ChREBP at Serine 568 by AMPK, Canis familiaris"
+xref: Reactome:REACT_86300 "pAMPK inactivates ACC2 inhibiting malonyl-CoA synthesis, Xenopus tropicalis"
+xref: Reactome:REACT_86799 "Phosphorylated AMPK phosphorylates TSC2, Sus scrofa"
+xref: Reactome:REACT_89177 "Phosphorylation of ChREBP at Serine 568 by AMPK, Gallus gallus"
+xref: Reactome:REACT_93047 "Phosphorylation of ChREBP at Serine 568 by AMPK, Taeniopygia guttata"
+xref: Reactome:REACT_93827 "pAMPK inactivates ACC2 inhibiting malonyl-CoA synthesis, Schizosaccharomyces pombe"
+xref: Reactome:REACT_95348 "pAMPK inactivates ACC2 inhibiting malonyl-CoA synthesis, Sus scrofa"
+xref: Reactome:REACT_96631 "Phosphorylated AMPK phosphorylates TSC2, Taeniopygia guttata"
+xref: Reactome:REACT_96928 "Phosphorylated AMPK phosphorylates TSC2, Gallus gallus"
+xref: Reactome:REACT_97108 "Phosphorylated AMPK phosphorylates TSC2, Canis familiaris"
+xref: Reactome:REACT_97911 "Phosphorylated AMPK phosphorylates TSC2, Drosophila melanogaster"
+xref: Reactome:REACT_98167 "pAMPK inactivates ACC2 inhibiting malonyl-CoA synthesis, Bos taurus"
+xref: Reactome:REACT_99741 "Phosphorylated AMPK phosphorylates TSC2, Mus musculus"
+is_a: GO:0004674 ! protein serine/threonine kinase activity
+is_a: GO:0060089 ! molecular transducer activity
+
+
+[Term]
+id: GO:0005072
+name: transforming growth factor beta receptor, cytoplasmic mediator activity
+namespace: molecular_function
+def: "Activity of any of the molecules that transmit the signal from a TGF-beta receptor through the cytoplasm to the nucleus." [GOC:hjd]
+comment: Note that this term and its children are useful for annotating proteins of the SMAD family.
+synonym: "TGF-beta receptor, cytoplasmic mediator activity" EXACT []
+synonym: "TGFbeta receptor, cytoplasmic mediator activity" EXACT []
+is_a: GO:0060089 ! molecular transducer activity
+
+[Term]
+id: GO:0003674
+name: molecular_function
+namespace: molecular_function
+alt_id: GO:0005554
+def: "A molecular process that can be carried out by the action of a single macromolecular machine, usually via direct physical interactions with other molecular entities. Function in this sense denotes an action, or activity, that a gene product (or a complex) performs. These actions are described from two distinct but related perspectives: (1) biochemical activity, and (2) role as a component in a larger system/process." [GOC:pdt]
+comment: Note that, in addition to forming the root of the molecular function ontology, this term is recommended for use for the annotation of gene products whose molecular function is unknown. When this term is used for annotation, it indicates that no information was available about the molecular function of the gene product annotated as of the date the annotation was made; the evidence code "no data" (ND), is used to indicate this. Despite its name, this is not a type of 'function' in the sense typically defined by upper ontologies such as Basic Formal Ontology (BFO). It is instead a BFO:process carried out by a single gene product or complex.
+subset: goslim_aspergillus
+subset: goslim_candida
+subset: goslim_chembl
+subset: goslim_generic
+subset: goslim_metagenomics
+subset: goslim_pir
+subset: goslim_plant
+subset: goslim_yeast
+synonym: "molecular function" EXACT []
+
+
+[Term]
+id: GO:0060089
+name: molecular transducer activity
+namespace: molecular_function
+def: "A compound molecular function in which an effector function is controlled by one or more regulatory components." [GOC:dos, GOC:pdt]
+subset: gocheck_do_not_manually_annotate
+subset: goslim_pir
+is_a: GO:0003674 ! molecular_function
+
+[Term]
+id: GO:0004690
+name: cyclic nucleotide-dependent protein kinase activity
+namespace: molecular_function
+def: "Catalysis of the reaction: ATP + a protein = ADP + a phosphoprotein. This reaction requires the presence of a cyclic nucleotide." [GOC:mah]
+xref: EC:2.7.11
+is_a: GO:0004674 ! protein serine/threonine kinase activity
+is_a: GO:0060089 ! molecular transducer activity
+
+[Term]
+id: GO:0009927
+name: histidine phosphotransfer kinase activity
+namespace: molecular_function
+def: "Serves as a phospho-His intermediate enabling the transfer of phospho group between a hybrid kinase and a response regulator." [PMID:11842140]
+comment: Added is_a molecular transducer activity see https://github.com/geneontology/go-ontology/issues/15611
+is_a: GO:0004672 ! protein kinase activity
+is_a: GO:0060089 ! molecular transducer activity
+
+[Term]
+id: GO:0017132
+name: cAMP-dependent guanyl-nucleotide exchange factor activity
+namespace: molecular_function
+def: "Stimulates the exchange of guanyl nucleotides associated with a GTPase, when activated by cAMP binding." [GOC:mah, PMID:9856955]
+is_a: GO:0005085 ! guanyl-nucleotide exchange factor activity
+is_a: GO:0060089 ! molecular transducer activity
+
+[Term]
+id: GO:0031992
+name: energy transducer activity
+namespace: molecular_function
+def: "The biological transducer activity that accepts energy and converts it to another form, often by transfer to another molecule within the cell." [GOC:go_curators]
+subset: goslim_pir
+synonym: "light harvesting activity" RELATED []
+synonym: "photon capture" RELATED []
+is_a: GO:0060089 ! molecular transducer activity
+
+[Term]
+id: GO:0038023
+name: signaling receptor activity
+namespace: molecular_function
+alt_id: GO:0004872
+alt_id: GO:0019041
+def: "Receiving a signal and transmitting it in the cell to initiate a change in cell activity. A signal is a physical entity or change in state that is used to transfer information in order to trigger a response." [GOC:bf, GOC:signaling]
+subset: goslim_agr
+subset: goslim_chembl
+subset: goslim_metagenomics
+subset: goslim_mouse
+subset: goslim_pir
+subset: goslim_plant
+synonym: "receptor activity" BROAD []
+synonym: "receptor activity involved in signal transduction" EXACT [GOC:bf]
+is_a: GO:0060089 ! molecular transducer activity
+created_by: rfoulger
+creation_date: 2011-08-01T02:45:27Z
+
+[Term]
+id: GO:0097472
+name: cyclin-dependent protein kinase activity
+namespace: molecular_function
+def: "Catalysis of the phosphorylation of an amino acid residue in a protein, usually according to the reaction: a protein + ATP = a phosphoprotein + ADP. This reaction requires the binding of a regulatory cyclin subunit and full activity requires stimulatory phosphorylation by a CDK-activating kinase (CAK)." [GOC:pr]
+is_a: GO:0004672 ! protein kinase activity
+is_a: GO:0060089 ! molecular transducer activity
+relationship: part_of GO:0051726 ! regulation of cell cycle
+created_by: paola
+creation_date: 2013-03-06T13:46:04Z

tests/tripal_chado/loaders/OBOImporterTest.php

@@ -12,35 +12,9 @@ class OBOImporterTest extends TripalTestCase {
 
 
   /**
-   * @group obo
-   * @ticket 525
-   */
-  public function test_PTO_loads_colon_issue() {
-
-    $name = 'core_test_PTO_mini';
-
-    $path = __DIR__ . '/../example_files/pto_colon.obo';
-
-    $this->load_obo($name, $path);
-
-    $exists = db_select('chado.cv', 'c')
-      ->fields('c', ['cv_id'])
-      ->condition('name', 'core_test_PTO_mini')
-      ->execute()
-      ->fetchField();
-    $this->assertNotFalse($exists);
-
-    //hte colon splitting issue: a new CV will created named fatty acid 18
-    $exists = db_select('chado.cv', 'c')
-      ->fields('c', ['cv_id'])
-      ->condition('name', 'fatty acid 18')
-      ->execute()
-      ->fetchField();
-    $this->assertFalse($exists);
-
-  }
-
-  /**
+   * Tests that the Goslim ontology loads using a remote URL.  Ensure subgroups
+   * load.
+   *
    * @group obo
    */
 
@@ -57,22 +31,22 @@ class OBOImporterTest extends TripalTestCase {
       ->execute()
       ->fetchField();
     $this->assertNotFalse($exists);
-    
+
     $exists = db_select('chado.cv', 'c')
       ->fields('c', ['cv_id'])
       ->condition('name', 'cellular_component')
       ->execute()
       ->fetchField();
     $this->assertNotFalse($exists);
-    
+
     $exists = db_select('chado.cv', 'c')
       ->fields('c', ['cv_id'])
       ->condition('name', 'molecular_function')
       ->execute()
       ->fetchField();
     $this->assertNotFalse($exists);
-    
-    
+
+
     $sql = "
       SELECT DISTINCT CVTP.value
       FROM {cvtermprop} CVTP
@@ -83,11 +57,18 @@ class OBOImporterTest extends TripalTestCase {
       WHERE CVTPT.name = 'Subgroup' and DB.name = 'GO' and CVTP.value = 'goslim_plant'
     ";
     $exists = chado_query($sql)->fetchField();
-    $this->assertNotFalse($exists);    
+    $this->assertNotFalse($exists);
   }
 
 
-  private function load_obo($name,$path){
+  /**
+   * @param $name - ontology name.  This goes in the tripal_cv_obo table.
+   * @param $path - path to the OBO.  this can be a file path or a URL.
+   *
+   * @throws \Exception
+   */
+
+  private function load_obo($name, $path) {
 
     $obo_id = db_select('public.tripal_cv_obo', 't')
       ->fields('t', ['obo_id'])
@@ -110,11 +91,11 @@ class OBOImporterTest extends TripalTestCase {
     $importer->prepareFiles();
     $importer->run();
 
-
   }
 
   /**
-   * @throws \Exception
+   * Ensure SO can be loaded.  Ensure that synonyms are set properly.
+   *
    * @group obo
    * @ticket 525
    */
@@ -122,8 +103,10 @@ class OBOImporterTest extends TripalTestCase {
 
     // step 1: drop the SO CV and CASCADE.
 
-    $result = chado_query("DELETE FROM {cv} WHERE name = 'sequence'");
-    $result = chado_query("DELETE FROM {db} WHERE name = 'SO'");
+    $result = chado_query("SET search_path to public, chado;
+    DELETE FROM {cv} WHERE name = 'sequence'");
+    $result = chado_query("SET search_path to public, chado;
+    DELETE FROM {db} WHERE name = 'SO'");
 
     // step 2: re-add SO.
     $name = 'Sequence Ontology';
@@ -131,7 +114,7 @@ class OBOImporterTest extends TripalTestCase {
 
     $this->load_obo($name, $path);
 
-   $sql = "SELECT CVT.name, CVTSYN.synonym
+    $sql = "SELECT CVT.name, CVTSYN.synonym
 FROM {cvterm} CVT
   INNER JOIN {dbxref} DBX on DBX.dbxref_id = CVT.dbxref_id
   INNER JOIN {db} on DB.db_id = DBX.db_id
@@ -140,13 +123,213 @@ WHERE DB.name = 'SO' and CVT.name = 'supercontig'
 ORDER BY DBX.accession";
 
 
-   $results = chado_query($sql)->fetchAll();
-  $result = $results[0];
+    $results = chado_query($sql)->fetchAll();
+    $result = $results[0];
+
+    $this->assertNotNull($result);
+    $this->assertNotEmpty($result);
+    $this->assertEquals("supercontig", $result->name);
+    $this->assertEquals("scaffold", $result->synonym);
+  }
+
+
+  /**
+   * Test simply that nodes are inserted.
+   ** @group obo
+   *
+   */
 
-   $this->assertNotNull($result);
-   $this->assertNotEmpty($result);
-   $this->assertEquals("scaffold", $result->synonym);
+  public function test_cvtermpath_cv_nodes_inserted() {
 
+    $name = 'path_test_mini';
+    $path = __DIR__ . '/../example_files/cvtermpath_test.obo';
+    $this->load_obo($name, $path);
+
+    //Check cvtermpath is sane.
+
+    $cv_name = 'cvtermpath_test';
+
+
+    $nodes = [
+      'node01',
+      'node02',
+      'node03',
+      'node04',
+      'node05',
+      'node06',
+      'node07',
+      'node08',
+      'node09',
+      'node10',
+      'node11',
+      'node12',
+      'node13',
+      'node14',
+    ];
+
+    //check nodes are inserted.
+    foreach ($nodes as $node) {
+
+      $query = db_select('chado.cvterm', 'cvt')
+        ->fields('cvt', ['cvterm_id', 'cv_id'])
+        ->condition('cvt.name', $node);
+      $query->join('chado.cv', 'cv', 'cvt.cv_id = cv.cv_id');
+      $query->condition('cv.name', $cv_name);
+      $result = $query->execute()
+        ->fetchObject();
+      $this->assertNotFalse($result);
+
+    }
   }
 
+
+  /**
+   * This data provider currently returns an array of data in the following manner:
+   * item[0] - the object node
+   * item[1] - an array containing a list of subject nodes, ie, nodes that claim they have a "is_a" relationship with this node.
+   *
+   * @return array
+   */
+  public function node_data_provider() {
+
+    $data = [
+      [
+        'node01',// object
+        [ //subjects
+          'node02',
+          'node03',
+          'node04',
+          'node05',
+          'node06',
+          'node07',
+          'node08',
+          'node09',
+          'node10',
+          'node11',
+          'node12',
+          'node13',
+          //   'node14',
+          //Node 14 is not connected!
+        ],
+      ],
+
+      [
+        'node04',
+        ['node09','node11'],
+      ],
+      [
+        'node11',
+        ['node13',],
+      ],
+    ];
+    return $data;
+
+  }
+
+  /**
+   * @group obo
+   * @dataProvider node_data_provider
+   */
+  public function test_cvtermpath_correct($object, $subjects) {
+
+
+    $name = 'path_test_mini';
+    $path = __DIR__ . '/../example_files/cvtermpath_test.obo';
+    $this->load_obo($name, $path);
+
+
+    $cv_name = 'cvtermpath_test';
+
+    $cv_id = chado_get_cv(['name' => $cv_name])->cv_id;
+
+    //populate cvtermpath
+    chado_update_cvtermpath($cv_id);
+
+
+    $query = db_select('chado.cvtermpath', 'cp');
+    $query->fields('cp', ['pathdistance']);
+    $query->condition('cp.cv_id', $cv_id);
+    $query->join('chado.cvterm', 'subject', 'cp.subject_id = subject.cvterm_id');
+    $query->join('chado.cvterm', 'object', 'cp.object_id = object.cvterm_id');
+    $query->condition('object.name', $object);
+
+    $query->fields('subject', ['name']);
+    // $delete_result = $query->execute()->fetchAll();
+
+
+    foreach ($subjects as $subject) {
+
+      $query_copy = clone $query;
+
+      $query_copy->condition('subject.name', $subject);
+      $results = $query_copy->execute()->fetchObject();
+
+      $this->assertNotFalse($results, "failed for {$object} as object and {$subject} as subject.");
+
+    }
+  }
+
+  /**
+   * ONLY node 1 should be root for test OBO.
+   *
+   * @group obo
+   */
+  public function test_cvtermpath_mview_root_terms_correct() {
+
+    $name = 'path_test_mini';
+    $path = __DIR__ . '/../example_files/cvtermpath_test.obo';
+    $this->load_obo($name, $path);
+
+    //populate mview
+    chado_populate_mview(chado_get_mview_id('cv_root_mview'));
+
+
+    $cv_name = 'cvtermpath_test';
+
+    $roots = db_select('chado.cv_root_mview', 't')
+      ->fields('t', ['name'])
+      ->condition('cv_name', $cv_name)
+      ->execute()
+      ->fetchAll();
+
+    $this->assertNotEmpty($roots);
+
+    $this->assertLessThan(2, count($roots));
+    $this->assertEquals("node01", $roots[0]->name);
+
+  }
+
+
+  /**
+   * ensure that new CV's aren't accidentally created when term names have
+   * colons in them.
+   *
+   * @group obo
+   * @ticket 525
+   */
+  public function test_PTO_loads_colon_issue() {
+
+    $name = 'core_test_PTO_mini';
+    $path = __DIR__ . '/../example_files/pto_colon.obo';
+
+    $this->load_obo($name, $path);
+
+    $exists = db_select('chado.cv', 'c')
+      ->fields('c', ['cv_id'])
+      ->condition('name', 'core_test_PTO_mini')
+      ->execute()
+      ->fetchField();
+    $this->assertNotFalse($exists);
+
+    //hte colon splitting issue: a new CV will created named fatty acid 18
+    $exists = db_select('chado.cv', 'c')
+      ->fields('c', ['cv_id'])
+      ->condition('name', 'fatty acid 18')
+      ->execute()
+      ->fetchField();
+    $this->assertFalse($exists);
+
+  }
+
+
 }