Add importer and integration test for structural variant data

core/src/test/java/org/mskcc/cbio/portal/scripts/TestImportStructuralVariantData.java

@@ -0,0 +1,150 @@
+/*
+ * Copyright (c) 2018 The Hyve B.V.
+ * This code is licensed under the GNU Affero General Public License (AGPL),
+ * version 3, or (at your option) any later version.
+ */
+
+/*
+ * This file is part of cBioPortal.
+ *
+ * cBioPortal is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU Affero General Public License as
+ * published by the Free Software Foundation, either version 3 of the
+ * License.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU Affero General Public License for more details.
+ *
+ * You should have received a copy of the GNU Affero General Public License
+ * along with this program. If not, see <http://www.gnu.org/licenses/>.
+*/
+
+/*
+ * @author Sander Tan
+*/
+
+package org.mskcc.cbio.portal.scripts;
+
+import org.junit.Before;
+import org.junit.Test;
+import org.junit.runner.RunWith;
+import org.mskcc.cbio.portal.dao.*;
+import org.mskcc.cbio.portal.model.*;
+import org.mskcc.cbio.portal.util.*;
+import org.springframework.test.context.ContextConfiguration;
+import org.springframework.test.context.junit4.SpringJUnit4ClassRunner;
+import org.springframework.test.context.transaction.TransactionConfiguration;
+import org.springframework.transaction.annotation.Transactional;
+import static org.junit.Assert.*;
+import java.io.*;
+import java.sql.*;
+
+/**
+ * Test class to test functionality of ImportStructralVariantData
+*/
+@SuppressWarnings("deprecation")
+@RunWith(SpringJUnit4ClassRunner.class)
+@ContextConfiguration(locations = { "classpath:/applicationContext-dao.xml" })
+@TransactionConfiguration(transactionManager = "transactionManager", defaultRollback = true)
+@Transactional
+public class TestImportStructuralVariantData{
+    int studyId;
+    int geneticProfileId;
+
+    /**
+     * Extracts StructuralVariant record from ResultSet.
+     * @param rs
+     * @return StructuralVariant record
+     * @throws SQLException
+     * @throws DaoException
+     */
+    private StructuralVariant extractStructuralVariant(ResultSet rs) throws SQLException, DaoException {
+        StructuralVariant structuralVariant = new StructuralVariant();
+        structuralVariant.setGeneticProfileId(rs.getInt("GENETIC_PROFILE_ID"));
+        structuralVariant.setSampleIdInternal(rs.getInt("SAMPLE_ID"));
+        structuralVariant.setSite1EntrezGeneId(rs.getLong("SITE1_ENTREZ_GENE_ID"));
+        structuralVariant.setSite1EnsemblTranscriptId(rs.getString("SITE1_ENSEMBL_TRANSCRIPT_ID"));
+        structuralVariant.setSite1Exon(rs.getInt("SITE1_EXON"));
+        structuralVariant.setSite1Chromosome(rs.getString("SITE1_CHROMOSOME"));
+        structuralVariant.setSite1Position(rs.getInt("SITE1_POSITION"));
+        structuralVariant.setSite1Description(rs.getString("SITE1_DESCRIPTION"));
+        structuralVariant.setSite2EntrezGeneId(rs.getLong("SITE2_ENTREZ_GENE_ID"));
+        structuralVariant.setSite2EnsemblTranscriptId(rs.getString("SITE2_ENSEMBL_TRANSCRIPT_ID"));
+        structuralVariant.setSite2Exon(rs.getInt("SITE2_EXON"));
+        structuralVariant.setSite2Chromosome(rs.getString("SITE2_CHROMOSOME"));
+        structuralVariant.setSite2Position(rs.getInt("SITE2_POSITION"));
+        structuralVariant.setSite2Description(rs.getString("SITE2_DESCRIPTION"));
+        structuralVariant.setSite2EffectOnFrame(rs.getString("SITE2_EFFECT_ON_FRAME"));
+        structuralVariant.setNcbiBuild(rs.getString("NCBI_BUILD"));
+        structuralVariant.setDnaSupport(rs.getString("DNA_SUPPORT"));
+        structuralVariant.setRnaSupport(rs.getString("RNA_SUPPORT"));
+        structuralVariant.setNormalReadCount(rs.getInt("NORMAL_READ_COUNT"));
+        structuralVariant.setTumorReadCount(rs.getInt("TUMOR_READ_COUNT"));
+        structuralVariant.setNormalVariantCount(rs.getInt("NORMAL_VARIANT_COUNT"));
+        structuralVariant.setTumorVariantCount(rs.getInt("TUMOR_VARIANT_COUNT"));
+        structuralVariant.setNormalPairedEndReadCount(rs.getInt("NORMAL_PAIRED_END_READ_COUNT"));
+        structuralVariant.setTumorPairedEndReadCount(rs.getInt("TUMOR_PAIRED_END_READ_COUNT"));
+        structuralVariant.setNormalSplitReadCount(rs.getInt("NORMAL_SPLIT_READ_COUNT"));
+        structuralVariant.setTumorSplitReadCount(rs.getInt("TUMOR_SPLIT_READ_COUNT"));
+        structuralVariant.setAnnotation(rs.getString("ANNOTATION"));
+        structuralVariant.setBreakpointType(rs.getString("BREAKPOINT_TYPE"));
+        structuralVariant.setCenter(rs.getString("CENTER"));
+        structuralVariant.setConnectionType(rs.getString("CONNECTION_TYPE"));
+        structuralVariant.setEventInfo(rs.getString("EVENT_INFO"));
+        structuralVariant.setVariantClass(rs.getString("CLASS"));
+        structuralVariant.setLength(rs.getInt("LENGTH"));
+        structuralVariant.setComments(rs.getString("COMMENTS"));
+        structuralVariant.setExternalAnnotation(rs.getString("EXTERNAL_ANNOTATION"));
+        structuralVariant.setDriverFilter(rs.getString("DRIVER_FILTER"));
+        structuralVariant.setDriverFilterAnn(rs.getString("DRIVER_FILTER_ANNOTATION"));
+        structuralVariant.setDriverTiersFilter(rs.getString("DRIVER_TIERS_FILTER"));
+        structuralVariant.setDriverTiersFilterAnn(rs.getString("DRIVER_TIERS_FILTER_ANNOTATION"));
+        return structuralVariant;
+    }
+
+    @Before
+    public void setUp() throws DaoException
+    {
+        studyId = DaoCancerStudy.getCancerStudyByStableId("study_tcga_pub").getInternalId();
+        geneticProfileId = DaoGeneticProfile.getGeneticProfileByStableId("study_tcga_pub_structural_variants").getGeneticProfileId();
+    }
+
+    @Test
+    public void testImportStructuralVariantData() throws DaoException, IOException {
+        ProgressMonitor.setConsoleMode(false);
+
+        // Load test structural variants
+        File file = new File("src/test/resources/data_structural_variants.txt");
+        ImportStructuralVariantData importer = new ImportStructuralVariantData(file, geneticProfileId, null);
+        importer.importData();
+        MySQLbulkLoader.flushAll();
+
+        // Retrieve all imported structural variants
+        Connection con = null;
+        PreparedStatement pstmt = null;
+        ResultSet rs = null;
+        try {
+            con = JdbcUtil.getDbConnection(DaoGeneset.class);
+            pstmt = con.prepareStatement("SELECT * FROM structural_variant");
+            rs = pstmt.executeQuery();
+
+            // Test first structural variant entry
+            rs.next();
+            StructuralVariant structuralVariant = extractStructuralVariant(rs);
+            assertEquals("KIAA1549-BRAF.K16B10.COSF509_2", structuralVariant.getSite2Description());
+
+            // Test second structural variant entry
+            rs.next();
+            structuralVariant = extractStructuralVariant(rs);
+            assertEquals("ENST00000318522", structuralVariant.getSite1EnsemblTranscriptId());
+        }
+        catch (SQLException e) {
+            throw new DaoException(e);
+        }
+        finally {
+            JdbcUtil.closeAll(DaoGeneset.class, con, pstmt, rs);
+        }
+    }
+}

core/src/test/java/org/mskcc/cbio/portal/scripts/TestIntegrationTest.java

@@ -67,7 +67,9 @@
 import org.springframework.test.context.transaction.TransactionConfiguration;
 import org.springframework.transaction.annotation.Transactional;
 import org.cbioportal.model.GenesetMolecularData;
+import org.cbioportal.model.StructuralVariant;
 import org.cbioportal.service.GenesetDataService;
+import org.cbioportal.service.StructuralVariantService;
 import org.mskcc.cbio.portal.dao.DaoGeneset;
 
 
@@ -130,7 +132,7 @@ public void testLoadStudyEs0() throws Throwable {
                 }
             }
             //check that there are only warnings for empty positions in fake data:
-            assertEquals(1, countWarnings);
+            assertEquals(0, countWarnings);
 
             //check that ALL data really got into DB correctly. In the spirit of integration tests,
             //we want to query via the same service layer as the one used by the web API here.
@@ -163,10 +165,22 @@ public void testLoadStudyEs0() throws Throwable {
             List<DBGeneticProfile> geneticProfiles = geneticProfileMapperLegacy.getGeneticProfiles(geneticProfileStableIds);
             assertEquals(geneticProfiles.size(), 0);
 
+            //===== Check STRUCTURAL VARIANT data ========
+            // 7 structural variant events are importerted, using 11 unique genes, over 2 samples
+            geneticProfileStableIds = new ArrayList<String>(Arrays.asList("study_es_0_structural_variants", "study_es_0_structural_variants"));
+            List<Integer> entrezGeneIds = new ArrayList<Integer>(Arrays.asList(57670, 673, 8031, 5979, 27436, 238, 7113, 2078, 1956, 238, 5774));
+            List<String> sampleIds = new ArrayList<String>(Arrays.asList("TCGA-A2-A04P-01", "TCGA-A1-A0SB-01"));
+
+            StructuralVariantService structuralVariantService = applicationContext.getBean(StructuralVariantService.class);
+            List<StructuralVariant> structuralVariants = structuralVariantService.fetchStructuralVariants(geneticProfileStableIds, entrezGeneIds, sampleIds);
+
+            // Check if all 7 structural variants are imported
+            assertEquals(7, structuralVariants.size());
+
             //===== Check CNA data ========
             geneticProfileStableIds = new ArrayList<String>();
             geneticProfileStableIds.add("study_es_0_gistic");
-            List<String> hugoGeneSymbols = new ArrayList<String>(Arrays.asList("ACAP3","AGRN","ATAD3A","ATAD3B","ATAD3C","AURKAIP1","ERCC5"));
+            ArrayList<String> hugoGeneSymbols = new ArrayList<String>(Arrays.asList("ACAP3","AGRN","ATAD3A","ATAD3B","ATAD3C","AURKAIP1","ERCC5"));
             List<DBProfileData> cnaProfileData = apiService.getGeneticProfileData(geneticProfileStableIds, hugoGeneSymbols, null, null);
             //there is data for 7 genes x 778 samples:
             assertEquals(7*778, cnaProfileData.size());

core/src/test/resources/integrationTestScript.xml

@@ -35,7 +35,7 @@
                     <value>--data</value>
                     <value>src/test/resources/genesets/study_es_0_genesets.gmt</value>
                     <value>--new-version</value>
-                    <value>1</value>
+                    <value>msigdb_6.1</value>
                 </list>
 
                 <!-- #2 Remove study if exists -->

core/src/test/scripts/test_data/study_es_0/data_fusions.txt

@@ -2,4 +2,3 @@ Hugo_Symbol	Entrez_Gene_Id	Center	Tumor_Sample_Barcode	Fusion	DNA_support	RNA_su
 FGFR3	2261	saturn	TEST-A2B8-01	TMPRSS2-ERG fusion	yes	yes	unknown	unknown
 ERBB2	2064	jupiter	TEST-A2FF-01	Fusion2	yes	yes	unknown	unknown
 ERBB2	2064	uranus	TCGA-GI-A2C8-01	ERBB2-TEST Fusion	yes	yes	unknown	frameshift
-TP53	7157	uranus	TCGA-GI-A2C8-01	TP53-TEST Fusion	yes	yes	unknown	frameshift

core/src/test/scripts/test_data/study_es_0/meta_gsva_pvalues.txt

@@ -6,4 +6,4 @@ source_stable_id: gsva_scores
 profile_name: Pvalues of GSVA scores on oncogenic signatures gene sets
 profile_description: Pvalues GSVA scores using mRNA expression data
 data_filename: data_gsva_pvalues.txt
-geneset_def_version: 1
+geneset_def_version: msigdb_6.1

core/src/test/scripts/test_data/study_es_0/meta_gsva_scores.txt

@@ -7,4 +7,4 @@ profile_name: GSVA scores on oncogenic signatures gene sets
 profile_description: GSVA scores using mRNA expression data
 data_filename: data_gsva_scores.txt
 show_profile_in_analysis_tab: true
-geneset_def_version: 1
+geneset_def_version: msigdb_6.1

persistence/persistence-mybatis/src/main/java/org/cbioportal/persistence/mybatis/StructuralVariantMyBatisRepository.java

@@ -40,6 +40,10 @@ public class StructuralVariantMyBatisRepository implements StructuralVariantRepo
     public List<StructuralVariant> fetchStructuralVariants(List<String> molecularProfileIds, 
             List<Integer> entrezGeneIds, List<String> sampleIds) {
 
+        if (molecularProfileIds.size() != sampleIds.size()) {
+            throw new RuntimeException("molecularProfileIds list and sampleIds list should be the same length.");
+        }
+
         return structuralVariantMapper.fetchStructuralVariants(molecularProfileIds,
                 entrezGeneIds, sampleIds);
     }