Hi there, I am trying to use the pipeline from the best practice to analysis some exome data. But I get confused on the step "MergeBamAlignment".
We used bwa to map the fastq file to the reference and then, generated unmapped bam file using picard. After we merged the aligned bam file (A) and unmapped file (B), we got a merged bam file(C). However, the merge file (C) has a smaller size than file A and B.
7.3G sample.aligned.bam(A) 5.7G sample.raw.bam (B) 4.8G sample.merged.bam(C) 6.3M sample.merged.bai
According to informations from the guidance that MergeBamAlignment takes metadata from a SAM or BAM file of unmapped reads (uBAM) and merges it with a SAM or BAM file containing alignment records for a subset of those reads, i thought the merged file would be the largest file in A, B, and C.(http://gatkforums.broadinstitute.org/gatk/discussion/5969/how-to-generate-a-bam-for-variant-discovery-long#step7)
Why the merged file is smaller than others?Is there anybody familiar with the usage of MergeBamAlignment?
Here is my command:
alignment
$bwa mem -t 12 \ -M \ -R "@RG\tID:ID001\tLB:sample001\tSM:sample001\tPL:ILLUMINA" \ $ref_dir/ucsc.hg19.fasta \ $data_dir/sample001/ID001_1.clean.fq.gz $data_dir/sample001/ID001_2.clean.fq.gz > $sam_dir/ID001.aligned.sam
sort sam
java -jar $picard/SortSam.jar \ INPUT=$sam_dir/ID001.aligned.sam \ OUTPUT=$bam_dir/ID001.aligned.bam \ SORT_ORDER=queryname \ VALIDATION_STRINGENCY=LENIENT \ CREATE_INDEX=true \ TMP_DIR=$tmp_dir
generate unmapped bam file
java -jar $picard/FastqToSam.jar \ FASTQ=$data_dir/sample001/ID001_1.clean.fq.gz \ FASTQ2=$data_dir/sample001/ID001_2.clean.fq.gz \ OUTPUT=$bam_dir/ID001.raw.bam \ READ_GROUP_NAME=ID001 \ SAMPLE_NAME=sample001 \ LIBRARY_NAME=ID001 \ PLATFORM=ILLUMINA \ SORT_ORDER=queryname \ TMP_DIR=$tmp_dir
merge two bam file
java -jar $picard/MergeBamAlignment.jar \ ALIGNED_BAM=$bam_dir/ID001.aligned.bam \ UNMAPPED_BAM=$bam_dir/ID001.raw.bam \ OUTPUT=$bam_dir/ID001.bam \ SORT_ORDER=coordinate \ R=$ref_dir/ucsc.hg19.fasta \ CREATE_INDEX=true \ ADD_MATE_CIGAR=true \ CLIP_ADAPTERS=false \ CLIP_OVERLAPPING_READS=true \ INCLUDE_SECONDARY_ALIGNMENTS=true \ MAX_INSERTIONS_OR_DELETIONS=-1 \ PRIMARY_ALIGNMENT_STRATEGY=MostDistant \ ATTRIBUTES_TO_RETAIN=XS \ PAIRED_RUN=TRUE \ TMP_DIR=$tmp_dir
I would be very grateful of any advice!
Zoe
