Picard SamtoFastq Extract Only Paired Reads

[fongchun]

Hi,

I am using the Picard SamtoFastq function to extract paired fastq files from a bam file. My input is like this:

[Thu Jul 19 00:13:45 PDT 2012] net.sf.picard.sam.SamToFastq INPUT=input.bam FASTQ=out.1.fastq SECOND_END_FASTQ=out.2.fastq VALIDATION_STRINGENCY=LENIENT OUTPUT_PER_RG=false RE_REVERSE=true INCLUDE_NON_PF_READS=false READ1_TRIM=0 READ2_TRIM=0 INCLUDE_NON_PRIMARY_ALIGNMENTS=false VERBOSITY=INFO QUIET=false COMPRESSION_LEVEL=5 MAX_RECORDS_IN_RAM=500000 CREATE_INDEX=false CREATE_MD5_FILE=false

What I've noticed is that the fastq files that are getting extracted or not the same size and actually differ in the number of sequences in each file. I think this has to do with the complication that the input bam file has both single and paired-end reads.

This is apparently when I try to run gsnap on the fastq files and it complains about how the read names do not match up.

Is there a way to force the Picard samtofastq function to extract only paired-end reads?

Thanks

[loituna]

Hi,

a little up
Was this problem solved? I am having the same issue :
I am using picard tool as followed :
java -jar SamToFastq.jar INPUT=<bamfile> FASTQ=outfile_1.fastq SECOND_END_FASTQ=outfile_2.fastq

And I also got file with different size for _1 (30G) and _2 (18G) file. The number of line is also ~ double for _1.

If I understood correctly, _2 file should be either same size as _1, if bam was constructed from paired end seq, or empty if not. I don't understand this in between output.

any clue?

thanks,

(if it can help, I am working on cghub, tcga data)

[swbarnes2]

I don't know about Picard tools, but you want to filter by the flags, and samtools view can do that.

samtools view -bf 1 my.bam > my_paired.bam

[loituna]

Hi,

Thanks for the tip!
And sorry, false alert, I resolved my problem. I think the picard tool was not the problem here but it was the original bam file.

thanks anyway