ReplaceReadGroups. Fixes BAM files without read group info. MissingReadGroup points people to this tool now. Please point users on the forum to this tool now. Will migrate to Picard.

git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5284 348d0f76-0448-11de-a6fe-93d51630548a

build.xml

@@ -367,6 +367,15 @@
           </manifest>
         </jar>
 
+        <jar jarfile="${dist.dir}/ModifyReadGroup.jar" whenmanifestonly="skip">
+          <fileset dir="${java.classes}">
+            <include name="**/tools/**/*.class" />
+          </fileset>
+          <manifest>
+            <attribute name="Main-Class" value="org.broadinstitute.sting.playground.tools.ModifyReadGroup" />
+          </manifest>
+        </jar>
+
         <jar jarfile="${dist.dir}/CompareBAMAlignments.jar" whenmanifestonly="skip">
           <fileset dir="${java.classes}">
             <include name="**/tools/**/*.class" />
@@ -450,6 +459,12 @@
             </manifest>
         </jar>
 
+        <jar jarfile="${dist.dir}/ModifyReadGroup.jar" update="true" whenmanifestonly="skip">
+            <manifest>
+                <attribute name="Class-Path" value="${jar.classpath}" />
+            </manifest>
+        </jar>
+
     </target>
 
     <target name="queue.manifests" depends="queue.jar, init.manifests" if="scala.include">

java/src/org/broadinstitute/sting/gatk/phonehome/GATKRunReport.java

@@ -345,10 +345,10 @@ public class GATKRunReport {
                 // Create an S3Object based on a file, with Content-Length set automatically and
                 // Content-Type set based on the file's extension (using the Mimetypes utility class)
                 S3Object fileObject = new S3Object(localFile);
-                logger.info("Created S3Object" + fileObject);
-                logger.info("Uploading " + localFile + " to AWS bucket");
+                //logger.info("Created S3Object" + fileObject);
+                //logger.info("Uploading " + localFile + " to AWS bucket");
                 S3Object s3Object = s3Service.putObject(REPORT_BUCKET_NAME, fileObject);
-                logger.info("Uploaded: " + s3Object);
+                logger.info("Uploaded to AWS: " + s3Object);
             } catch ( S3ServiceException e ) {
                 exceptDuringRunReport("S3 exception occurred", e);
             } catch ( NoSuchAlgorithmException e ) {

java/src/org/broadinstitute/sting/oneoffprojects/tools/CompareBAMAlignments.java

@@ -61,10 +61,12 @@ public class CompareBAMAlignments extends CommandLineProgram {
 
                         if ( ! read1.getReadName().equals(read.getReadName()) )
                             bad(read1, read, "Names not equal");
-                        if ( read1.getAlignmentStart() != read.getAlignmentStart() )
-                            bad(read1, read, "Alignment starts not equal");
-                        if ( ! read1.getCigarString().equals(read.getCigarString()) )
-                            bad(read1, read, "Unequal CIGAR strings");
+                        else {
+                            if ( read1.getAlignmentStart() != read.getAlignmentStart() )
+                                bad(read1, read, "Alignment starts not equal");
+                            if ( ! read1.getCigarString().equals(read.getCigarString()) )
+                                bad(read1, read, "Unequal CIGAR strings");
+                        }
                     }
                 }
                 counter++;
@@ -79,8 +81,8 @@ public class CompareBAMAlignments extends CommandLineProgram {
 
     private void bad(SAMRecord read1, SAMRecord read2, String msg) {
         System.out.printf("%nBAD: %s%n", msg);
-        System.out.printf("  read1:  %s %s %s%n", read1.getReadName(), read1.getAlignmentStart(), read1.getCigarString());
-        System.out.printf("  read2:  %s %s %s%n", read2.getReadName(), read2.getAlignmentStart(), read2.getCigarString());
+        System.out.printf("  read1:  %s %s %s %s%n", read1.getReadName(), read1.getAlignmentStart(), read1.getCigarString(), read1.getInferredInsertSize());
+        System.out.printf("  read2:  %s %s %s %s%n", read2.getReadName(), read2.getAlignmentStart(), read2.getCigarString(), read2.getInferredInsertSize());
         // System.exit(1);
     }
 

java/src/org/broadinstitute/sting/playground/tools/ReplaceReadGroups.java

@@ -0,0 +1,78 @@
+package org.broadinstitute.sting.playground.tools;
+
+import net.sf.picard.cmdline.CommandLineProgram;
+import net.sf.picard.cmdline.Option;
+import net.sf.picard.cmdline.Usage;
+import net.sf.samtools.*;
+
+import java.io.File;
+import java.util.Arrays;
+
+/**
+ * User: mdepristo
+ *
+ * Replaces read groups in a BAM file
+ */
+public class ReplaceReadGroups extends CommandLineProgram {
+    @Usage(programVersion="1.0") public String USAGE = "Creates a new read group, and assigns all reads from the I BAM file to this read group in the O BAM";
+    @Option(shortName="I", doc="Input file (bam or sam).", optional=false)
+    public File IN = null;
+    @Option(shortName="O",doc="Output file (bam or sam).", optional=false)
+    public File OUT = null;
+
+    @Option(shortName="ID",doc="Read Group ID", optional=false)
+    public String RGID = null;
+
+    @Option(shortName="LB",doc="Read Group Library", optional=false)
+    public String RGLB = null;
+
+    @Option(shortName="PL",doc="Read Group platform", optional=false)
+    public String RGPL = null;
+
+    @Option(shortName="SM",doc="Read Group sample", optional=false)
+    public String RGSM = null;
+
+    private static final String RGFIELD = "RG"; // todo -- use binary tag that's private in picard
+
+    // todo -- is it worth supporting these fields?
+    // CN Name of sequencing center producing the read.
+    // DS Description.
+    // DT Date the run was produced (ISO8601 date or date/time).
+    // PU Platform unit (e.g. flowcell-barcode.lane for Illumina or slide for SOLiD). Unique identier.
+
+    /** Required main method implementation. */
+    public static void main(final String[] argv) {
+        System.exit(new ReplaceReadGroups().instanceMain(argv));
+    }
+
+    protected int doWork() {
+        SAMFileReader inReader = new SAMFileReader(IN);
+
+        // create the read group we'll be using
+        SAMReadGroupRecord rg = new SAMReadGroupRecord(RGID);
+        rg.setLibrary(RGLB);
+        rg.setPlatform(RGPL);
+        rg.setSample(RGSM);
+        System.out.printf("Created read group ID=%s PL=%s LB=%s SM=%s%n", rg.getId(), rg.getPlatform(), rg.getLibrary(), rg.getSample());
+
+        // create the new header and output file
+        SAMFileHeader outHeader = inReader.getFileHeader().clone();
+        outHeader.setReadGroups(Arrays.asList(rg));
+        SAMFileWriter outWriter = new SAMFileWriterFactory().makeSAMOrBAMWriter(outHeader, true, OUT) ;
+
+        //
+        // write the reads in contig order
+        //
+        for ( SAMRecord read : inReader ) {
+            read.setAttribute(RGFIELD, rg.getId());
+            outWriter.addAlignment(read);
+        }
+
+        // cleanup
+        inReader.close();
+        outWriter.close();
+        return 0;
+    }
+}
+
+

java/src/org/broadinstitute/sting/utils/exceptions/UserException.java

@@ -124,7 +124,7 @@ public class UserException extends ReviewedStingException {
 
     public static class ReadMissingReadGroup extends MalformedBam {
         public ReadMissingReadGroup(SAMRecord read) {
-            super(read, String.format("Read %s is either missing the read group or its read group is not defined in the BAM header, both of which are required by the GATK", read.getReadName()));
+            super(read, String.format("Read %s is either missing the read group or its read group is not defined in the BAM header, both of which are required by the GATK.  Please use http://www.broadinstitute.org/gsa/wiki/index.php/ReplaceReadGroups to fix this problem", read.getReadName()));
         }
     }