modified Bam to bed algorithm

[jarwulf]

I changed the Bam to Bed code I found here http://sourceforge.net/apps/mediawik...s_.28Picard.29 removing the commandline stuff to make it into simple conversion function that takes and passes a file back. Anything wrong with it? I'm new to Java so it may be riddled with errors.

Calling code

       //ALT if BAM 
        if (bamButton.isSelected())
        	File selectedFileBamHolder = BamToBed.doWork(selectedFile); //selectedFileBamHolder holds BAM since dunno if selectedFile could equal itself being processed by doWork
        	File selectedFile = selectedFileBamHolder;
        
        //ALT if not BAM just continue with other stuff

Conversion code

package gui;


import net.sf.picard.io.IoUtil;
import net.sf.samtools.*;

import java.io.File;
import java.util.Iterator;

/**
 * method for converting bam or sam files to bed files.
 */
public class BamToBed {

    File convertedBAM; //ALT File to pass back
    
    /** Whether the user provided sequence, start, and end args on the command line */
    protected boolean rangeArgsProvided = false;

    /** This method contains the main logic of the application */
    protected File doWork(File INPUT) { //ALT takes the (BAM) file inputted and processes it
        IoUtil.assertFileIsReadable(INPUT);

        final SAMFileReader reader = new SAMFileReader(INPUT);

        Iterator<SAMRecord> iterator = null;
        if(!rangeArgsProvided )
        {
            iterator = reader.iterator();
        }
        else
        {
            iterator = reader.queryOverlapping(SEQUENCE, START, END);
        }

        
        while (iterator.hasNext()) {
            final SAMRecord record = iterator.next();
            if (record.getReadUnmappedFlag()) {
                continue;
            }

            //Output is redirected from System.out to a File to be passed back to calling function
            FileWriter fstream = new FileWriter(convertedBAM);
            BufferedWriter out = new BufferedWriter(fstream);
            out.write(record.getReferenceName() + "\t" +
                    (record.getAlignmentStart() - 1) + "\t" + //subtract 1 to shift from one-based to zero-based
                    (record.getAlignmentEnd() - 1 + 1) + "\t" + //subtract 1 to shift from one-based to zero-based, and
                                                                // then add 1 to shift from inclusive to exclusive
                    record.getReadName() + "\t" +
                    record.getMappingQuality() + "\t" +
                    (record.getReadNegativeStrandFlag()? "-": "+") );
            out.close();
               
            
        }
        reader.close();

        return convertedBAM;
    }


}

[sdriscoll]

It seems like you should have to deal with the cigar notation stuff in there. Spliced alignments, deletions, etc. bed files have a way of annotating those as well. Either that or sometimes a single BAM line will have to be split into multiple BED lines.