Merge pull request #39 from cBioPortal/inc-seg

(7/7) RFC79: Implement incremental upload of CNA segmented data

scripts/importer/cbioportal_common.py

@@ -383,6 +383,7 @@ class MetaFileTypes(object):
     MetaFileTypes.TIMELINE,
     MetaFileTypes.GENE_PANEL_MATRIX,
     MetaFileTypes.STRUCTURAL_VARIANT,
+    MetaFileTypes.SEG,
 ]
 
 IMPORTER_CLASSNAME_BY_META_TYPE = {

src/main/java/org/mskcc/cbio/portal/dao/DaoClinicalData.java

@@ -370,7 +370,6 @@ public static List<ClinicalData> getSampleData(int cancerStudyId, Collection<Str
     public static void removeSampleAttributesData(int sampleInternalId) throws DaoException {
         Connection con = null;
         PreparedStatement pstmt = null;
-        ResultSet rs = null;
         try {
             con = JdbcUtil.getDbConnection(DaoClinicalData.class);
             pstmt = con.prepareStatement(SAMPLE_ATTRIBUTES_DELETE);
@@ -381,7 +380,31 @@ public static void removeSampleAttributesData(int sampleInternalId) throws DaoEx
             throw new DaoException(e);
         }
         finally {
-            JdbcUtil.closeAll(DaoClinicalData.class, con, pstmt, rs);
+            JdbcUtil.closeAll(DaoClinicalData.class, con, pstmt, null);
+        }
+    }
+
+    public static void removeSampleAttributesData(Set<Integer> sampleInternalIds, String attrId) throws DaoException {
+        Connection con = null;
+        PreparedStatement pstmt = null;
+        try {
+            con = JdbcUtil.getDbConnection(DaoClinicalData.class);
+            pstmt = con.prepareStatement("DELETE FROM " + SAMPLE_ATTRIBUTES_TABLE
+                    + " WHERE `ATTR_ID` = ? AND `INTERNAL_ID` IN ("
+                    + String.join(",", Collections.nCopies(sampleInternalIds.size(), "?"))
+                    + ")");
+            int parameterIndex = 1;
+            pstmt.setString(parameterIndex++, attrId);
+            for (Integer sampleInternalId : sampleInternalIds) {
+                pstmt.setInt(parameterIndex++, sampleInternalId);
+            }
+            pstmt.executeUpdate();
+        }
+        catch (SQLException e) {
+            throw new DaoException(e);
+        }
+        finally {
+            JdbcUtil.closeAll(DaoClinicalData.class, con, pstmt, null);
         }
     }
 

src/main/java/org/mskcc/cbio/portal/dao/DaoCopyNumberSegment.java

@@ -67,7 +67,7 @@ public static int addCopyNumberSegment(CopyNumberSegment seg) throws DaoExceptio
         }
     }
 
-    public static void createFractionGenomeAlteredClinicalData(int cancerStudyId) throws DaoException {
+    public static void createFractionGenomeAlteredClinicalData(int cancerStudyId, Set<Integer> sampleIds, boolean updateMode) throws DaoException {
         Connection con = null;
         PreparedStatement pstmt = null;
         ResultSet rs = null;
@@ -80,8 +80,15 @@ public static void createFractionGenomeAlteredClinicalData(int cancerStudyId) th
                     "AS c2 WHERE c2.`CANCER_STUDY_ID` = c1.`CANCER_STUDY_ID` AND c2.`SAMPLE_ID` = c1.`SAMPLE_ID` AND " + 
                     "ABS(c2.`SEGMENT_MEAN`) >= 0.2) / SUM(`END`-`START`)) AS `VALUE` FROM `copy_number_seg` AS c1 , `cancer_study` " +
                     "WHERE c1.`CANCER_STUDY_ID` = cancer_study.`CANCER_STUDY_ID` AND cancer_study.`CANCER_STUDY_ID`=? " +
-                    "GROUP BY cancer_study.`CANCER_STUDY_ID` , `SAMPLE_ID` HAVING SUM(`END`-`START`) > 0;");
-            pstmt.setInt(1, cancerStudyId);
+                            (sampleIds == null ? "" : ("AND `SAMPLE_ID` IN ("+ String.join(",", Collections.nCopies(sampleIds.size(), "?")) + ") "))
+                    +"GROUP BY cancer_study.`CANCER_STUDY_ID` , `SAMPLE_ID` HAVING SUM(`END`-`START`) > 0;");
+            int parameterIndex = 1;
+            pstmt.setInt(parameterIndex++, cancerStudyId);
+            if (sampleIds != null) {
+                for (Integer sampleId : sampleIds) {
+                    pstmt.setInt(parameterIndex++, sampleId);
+                }
+            }
             Map<Integer, String> fractionGenomeAltereds = new HashMap<Integer, String>();
             rs = pstmt.executeQuery();
             while (rs.next()) {
@@ -94,7 +101,10 @@ public static void createFractionGenomeAlteredClinicalData(int cancerStudyId) th
                     false, "20", cancerStudyId);
                 DaoClinicalAttributeMeta.addDatum(attr);
             }
-
+
+            if (updateMode) {
+                DaoClinicalData.removeSampleAttributesData(fractionGenomeAltereds.keySet(), FRACTION_GENOME_ALTERED_ATTR_ID);
+            }
             for (Map.Entry<Integer, String> fractionGenomeAltered : fractionGenomeAltereds.entrySet()) {
                 DaoClinicalData.addSampleDatum(fractionGenomeAltered.getKey(), FRACTION_GENOME_ALTERED_ATTR_ID, fractionGenomeAltered.getValue());
             }
@@ -283,4 +293,27 @@ public static boolean segmentDataExistForSample(int cancerStudyId, int sampleId)
             JdbcUtil.closeAll(DaoCopyNumberSegment.class, con, pstmt, rs);
         }
     }
+
+    public static void  deleteSegmentDataForSamples(int cancerStudyId, Set<Integer> sampleIds) throws DaoException {
+        Connection con = null;
+        PreparedStatement pstmt = null;
+        ResultSet rs = null;
+        try {
+            con = JdbcUtil.getDbConnection(DaoCopyNumberSegment.class);
+            pstmt = con.prepareStatement("DELETE FROM `copy_number_seg`" +
+                    " WHERE `CANCER_STUDY_ID`= ?" +
+                    " AND `SAMPLE_ID` IN (" + String.join(",", Collections.nCopies(sampleIds.size(), "?"))
+                    + ")");
+            int parameterIndex = 1;
+            pstmt.setInt(parameterIndex++, cancerStudyId);
+            for (Integer sampleId : sampleIds) {
+                pstmt.setInt(parameterIndex++, sampleId);
+            }
+            pstmt.executeUpdate();
+        } catch (SQLException e) {
+            throw new DaoException(e);
+        } finally {
+            JdbcUtil.closeAll(DaoCopyNumberSegment.class, con, pstmt, rs);
+        }
+    }
 }

src/main/java/org/mskcc/cbio/portal/dao/DaoCopyNumberSegmentFile.java

@@ -65,7 +65,7 @@ public static int addCopyNumberSegmentFile(CopyNumberSegmentFile copySegFile) th
         } catch (SQLException e) {
             throw new DaoException(e);
         } finally {
-            JdbcUtil.closeAll(DaoCopyNumberSegment.class, con, pstmt, rs);
+            JdbcUtil.closeAll(DaoCopyNumberSegmentFile.class, con, pstmt, rs);
         }
     }
 
@@ -86,6 +86,9 @@ public static CopyNumberSegmentFile getCopyNumberSegmentFile(int cancerStudyId)
                 cnsf.referenceGenomeId = CopyNumberSegmentFile.ReferenceGenomeId.valueOf(rs.getString("REFERENCE_GENOME_ID"));
                 cnsf.description = rs.getString("DESCRIPTION");
                 cnsf.filename = rs.getString("FILENAME");
+                if (rs.next()) {
+                    throw new SQLException("More than one row was returned.");
+                }
                 return cnsf;
             }
             return null;

src/main/java/org/mskcc/cbio/portal/scripts/ImportCopyNumberSegmentData.java

@@ -55,7 +55,9 @@
 import java.io.FileReader;
 import java.io.IOException;
 import java.math.BigDecimal;
+import java.util.HashSet;
 import java.util.Properties;
+import java.util.Set;
 
 /**
  * Import Segment data into database.
@@ -64,14 +66,17 @@
 public class ImportCopyNumberSegmentData extends ConsoleRunnable {
 
     private int entriesSkipped;
-
+    private boolean updateMode;
+    private Set<Integer> processedSampleIds;
+
     private void importData(File file, int cancerStudyId) throws IOException, DaoException {
         MySQLbulkLoader.bulkLoadOn();
         FileReader reader = new FileReader(file);
         BufferedReader buf = new BufferedReader(reader);
         try {
             String line = buf.readLine(); // skip header line
             long segId = DaoCopyNumberSegment.getLargestId();
+            processedSampleIds = new HashSet<>();
             while ((line=buf.readLine()) != null) {
                 ProgressMonitor.incrementCurValue();
                 ConsoleUtil.showProgress();
@@ -81,8 +86,7 @@ private void importData(File file, int cancerStudyId) throws IOException, DaoExc
                     System.err.println("wrong format: "+line);
                 }
 
-                CancerStudy cancerStudy = DaoCancerStudy.getCancerStudyByInternalId(cancerStudyId);
-                String chrom = strs[1].trim(); 
+                String chrom = strs[1].trim();
                 //validate in same way as GistitReader:
                 ValidationUtils.validateChromosome(chrom);
 
@@ -112,6 +116,10 @@ private void importData(File file, int cancerStudyId) throws IOException, DaoExc
                 CopyNumberSegment cns = new CopyNumberSegment(cancerStudyId, s.getInternalId(), chrom, start, end, numProbes, segMean);
                 cns.setSegId(++segId);
                 DaoCopyNumberSegment.addCopyNumberSegment(cns);
+                processedSampleIds.add(s.getInternalId());
+            }
+            if (updateMode) {
+                DaoCopyNumberSegment.deleteSegmentDataForSamples(cancerStudyId, processedSampleIds);
             }
             MySQLbulkLoader.flushAll();
         }
@@ -127,6 +135,7 @@ public void run() {
             OptionSet options = ConsoleUtil.parseStandardDataAndMetaOptions(args, description, true);
             String dataFile = (String) options.valueOf("data");
             File descriptorFile = new File((String) options.valueOf("meta"));
+            updateMode = options.has("overwrite-existing");
 
             Properties properties = new Properties();
             properties.load(new FileInputStream(descriptorFile));
@@ -135,13 +144,13 @@ public void run() {
 
             CancerStudy cancerStudy = getCancerStudy(properties);
 
-            if (segmentDataExistsForCancerStudy(cancerStudy)) {
+            if (!updateMode && segmentDataExistsForCancerStudy(cancerStudy)) {
                  throw new IllegalArgumentException("Seg data for cancer study " + cancerStudy.getCancerStudyStableId() + " has already been imported: " + dataFile);
             }
 
             importCopyNumberSegmentFileMetadata(cancerStudy, properties);
             importCopyNumberSegmentFileData(cancerStudy, dataFile);
-            DaoCopyNumberSegment.createFractionGenomeAlteredClinicalData(cancerStudy.getInternalId());
+            DaoCopyNumberSegment.createFractionGenomeAlteredClinicalData(cancerStudy.getInternalId(), processedSampleIds, updateMode);
             if( MySQLbulkLoader.isBulkLoad()) {
                 MySQLbulkLoader.flushAll();
             }
@@ -164,7 +173,7 @@ private static boolean segmentDataExistsForCancerStudy(CancerStudy cancerStudy)
         return (DaoCopyNumberSegment.segmentDataExistForCancerStudy(cancerStudy.getInternalId()));
     }
 
-    private static void importCopyNumberSegmentFileMetadata(CancerStudy cancerStudy, Properties properties) throws DaoException {
+    private void importCopyNumberSegmentFileMetadata(CancerStudy cancerStudy, Properties properties) throws DaoException {
         CopyNumberSegmentFile copyNumSegFile = new CopyNumberSegmentFile();
         copyNumSegFile.cancerStudyId = cancerStudy.getInternalId();
         String referenceGenomeId = properties.getProperty("reference_genome_id").trim();
@@ -179,7 +188,18 @@ private static void importCopyNumberSegmentFileMetadata(CancerStudy cancerStudy,
         copyNumSegFile.referenceGenomeId = getRefGenId(referenceGenomeId); 
         copyNumSegFile.description = properties.getProperty("description").trim();
         copyNumSegFile.filename = properties.getProperty("data_filename").trim();
-        DaoCopyNumberSegmentFile.addCopyNumberSegmentFile(copyNumSegFile);
+        CopyNumberSegmentFile storedCopyNumSegFile = DaoCopyNumberSegmentFile.getCopyNumberSegmentFile(cancerStudy.getInternalId());
+        if (updateMode && storedCopyNumSegFile != null) {
+            if (storedCopyNumSegFile.referenceGenomeId != copyNumSegFile.referenceGenomeId) {
+                throw new IllegalStateException("You are trying to upload "
+                        + copyNumSegFile.referenceGenomeId
+                        + " reference genome data into "
+                        + storedCopyNumSegFile.referenceGenomeId
+                        + " reference genome data.");
+            }
+        } else {
+            DaoCopyNumberSegmentFile.addCopyNumberSegmentFile(copyNumSegFile);
+        }
     }
 
     private void importCopyNumberSegmentFileData(CancerStudy cancerStudy, String dataFilename) throws IOException, DaoException {

src/main/java/org/mskcc/cbio/portal/scripts/ImportTimelineData.java

@@ -135,8 +135,13 @@ public void run() {
         try {
 		    String description = "Import 'timeline' data";
 
-		    OptionSet options = ConsoleUtil.parseStandardDataAndMetaOptions(args, description, false);
-		    String dataFile = (String) options.valueOf("data");
+		    OptionSet options = ConsoleUtil.parseStandardDataAndMetaOptions(args, description, true);
+			if (options.has("loadMode") && !"bulkLoad".equals(options.valueOf("loadMode"))) {
+				throw new UnsupportedOperationException("This loader supports bulkLoad load mode only, but "
+						+ options.valueOf("loadMode")
+						+ " has been supplied.");
+			}
+			String dataFile = (String) options.valueOf("data");
 		    File descriptorFile = new File((String) options.valueOf("meta"));
 			boolean overwriteExisting = options.has("overwrite-existing");
 

src/main/java/org/mskcc/cbio/portal/util/ConsoleUtil.java

@@ -178,10 +178,6 @@ public static OptionSet parseStandardDataAndMetaOptions(String[] args, String de
 							"Error: unknown loadMode action:  " + actionArg);
 				}
 			}
-			else {
-				throw new UsageException(progName, description, parser,
-						"Error: 'loadMode' argument required.");
-			}
 		}
 		return options;
 	}

src/test/java/org/mskcc/cbio/portal/integrationTest/incremental/TestIncrementalCopyNumberSegmentDataImport.java

@@ -0,0 +1,121 @@
+/*
+ * 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/>.
+*/
+
+package org.mskcc.cbio.portal.integrationTest.incremental;
+
+import org.junit.Before;
+import org.junit.Test;
+import org.junit.runner.RunWith;
+import org.mskcc.cbio.portal.dao.DaoCancerStudy;
+import org.mskcc.cbio.portal.dao.DaoClinicalData;
+import org.mskcc.cbio.portal.dao.DaoCopyNumberSegment;
+import org.mskcc.cbio.portal.dao.DaoCopyNumberSegmentFile;
+import org.mskcc.cbio.portal.dao.DaoException;
+import org.mskcc.cbio.portal.dao.DaoSample;
+import org.mskcc.cbio.portal.dao.MySQLbulkLoader;
+import org.mskcc.cbio.portal.model.CancerStudy;
+import org.mskcc.cbio.portal.model.ClinicalData;
+import org.mskcc.cbio.portal.model.CopyNumberSegment;
+import org.mskcc.cbio.portal.model.CopyNumberSegmentFile;
+import org.mskcc.cbio.portal.model.Sample;
+import org.mskcc.cbio.portal.scripts.ImportCopyNumberSegmentData;
+import org.springframework.test.annotation.Rollback;
+import org.springframework.test.context.ContextConfiguration;
+import org.springframework.test.context.junit4.SpringJUnit4ClassRunner;
+import org.springframework.transaction.annotation.Transactional;
+
+import java.io.File;
+import java.util.List;
+import java.util.Set;
+
+import static org.junit.Assert.assertEquals;
+import static org.junit.Assert.assertNotNull;
+
+/**
+ * Tests Incremental Import of CNA segmented data.
+ *
+ * @author Ruslan Forostianov
+ * @author Pieter Lukasse
+ */
+@RunWith(SpringJUnit4ClassRunner.class)
+@ContextConfiguration(locations = { "classpath:/applicationContext-dao.xml" })
+@Rollback
+@Transactional
+public class TestIncrementalCopyNumberSegmentDataImport {
+
+    /**
+     * Test incremental upload of CNA SEG data
+     */
+	@Test
+    public void testIncrementalUpload() throws DaoException {
+        String segSampleId = "TCGA-A1-A0SE-01";
+        Sample segDataSample = DaoSample.getSampleByCancerStudyAndSampleId(cancerStudy.getInternalId(), segSampleId);
+
+        CopyNumberSegmentFile copyNumberSegmentFile = new CopyNumberSegmentFile();
+        copyNumberSegmentFile.cancerStudyId = cancerStudy.getInternalId();
+        copyNumberSegmentFile.referenceGenomeId = CopyNumberSegmentFile.ReferenceGenomeId.hg19;
+        copyNumberSegmentFile.segFileId = 1;
+        copyNumberSegmentFile.filename = "test_file.seg";
+        copyNumberSegmentFile.description = "test seg file description";
+        DaoCopyNumberSegmentFile.addCopyNumberSegmentFile(copyNumberSegmentFile);
+        DaoClinicalData.addSampleDatum(segDataSample.getInternalId(), "FRACTION_GENOME_ALTERED", "TEST");
+        MySQLbulkLoader.bulkLoadOn();
+        CopyNumberSegment copyNumberSegment = new CopyNumberSegment(
+                cancerStudy.getInternalId(),
+                segDataSample.getInternalId(),
+                "1",
+                3218610,
+                95674710,
+                100,
+                0.01);
+        copyNumberSegment.setSegId(1L);
+        DaoCopyNumberSegment.addCopyNumberSegment(copyNumberSegment);
+        MySQLbulkLoader.flushAll();
+
+        File dataFolder = new File("src/test/resources/incremental/copy_number_alteration/");
+        File metaFile = new File(dataFolder, "meta_cna_seg.txt");
+        File dataFile = new File(dataFolder, "data_cna.seg");
+
+        ImportCopyNumberSegmentData importCnaSegData = new ImportCopyNumberSegmentData(new String[] {
+                "--loadMode", "bulkLoad",
+                "--meta", metaFile.getAbsolutePath(),
+                "--data", dataFile.getAbsolutePath(),
+                "--overwrite-existing",
+        });
+        importCnaSegData.run();
+
+        CopyNumberSegmentFile fetchedCopyNumberSegmentFile = DaoCopyNumberSegmentFile.getCopyNumberSegmentFile(cancerStudy.getInternalId());
+        assertNotNull(fetchedCopyNumberSegmentFile);
+        assertEquals("test_file.seg", fetchedCopyNumberSegmentFile.filename);
+        List<CopyNumberSegment> cnaSegments = DaoCopyNumberSegment
+                .getSegmentForASample(segDataSample.getInternalId(), cancerStudy.getInternalId());
+        assertEquals(9, cnaSegments.size());
+        List<ClinicalData> clinicalData = DaoClinicalData.getSampleData(cancerStudy.getInternalId(), Set.of(segSampleId));
+        ClinicalData fractionGenomeAltered = clinicalData.stream()
+                .filter(cd -> "FRACTION_GENOME_ALTERED".equals(cd.getAttrId())).findFirst().get();
+        assertEquals("0.0000", fractionGenomeAltered.getAttrVal());
+    }
+
+    public static final String STUDY_ID = "study_tcga_pub";
+    private CancerStudy cancerStudy;
+
+    @Before
+    public void setUp() throws DaoException {
+        cancerStudy = DaoCancerStudy.getCancerStudyByStableId(STUDY_ID);
+    }
+
+}

src/test/resources/incremental/copy_number_alteration/data_cna.seg

@@ -0,0 +1,10 @@
+ID	chrom	loc.start	loc.end	num.mark	seg.mean
+TCGA-A1-A0SE-01	1	3218610	95674710	53225	0.0055
+TCGA-A1-A0SE-01	1	95676511	95676518	2	-1.6636
+TCGA-A1-A0SE-01	1	95680124	167057183	24886	0.0053
+TCGA-A1-A0SE-01	1	167057495	167059336	3	-1.0999
+TCGA-A1-A0SE-01	1	167059760	181602002	9213	-8e-04
+TCGA-A1-A0SE-01	1	181603120	181609567	6	-1.2009
+TCGA-A1-A0SE-01	1	181610685	201473647	12002	0.0055
+TCGA-A1-A0SE-01	1	201474400	201474544	2	-1.4235
+TCGA-A1-A0SE-01	1	201475220	247813706	29781	-4e-04