FastqToSam Picard tool error

[incognitos]

Hi all,

I've been running into a problem with the FastqToSam Picard tool:

My run command is: java -Xmx4g -jar /n/sw/picard-tools-1.98//FastqToSam.jar FASTQ=file.R1.fastq.gz FASTQ2=file.R2.fastq.gz SAMPLE_NAME=2a OUTPUT=text.bam

The error I get is: Exception in thread "main" net.sf.picard.PicardException: Missing sequence line at line 54573 in fastq /path/file.R2.fastq.gz

Anyone know what's going on?

[mastal]

Is there something wrong with the format of the fastq file at the line indicated by the error message?

[incognitos]

Yes, I just checked that.
It seems that there is no read2 sequence on that line, although read1 exists.

Should I delete those lines in the R1 and R2 fastq files?

PS: This is what R2 looks like around the line that is giving the error:
////////<//E//////<///<//<//E////<//<///////////////////<///
@NS500422:212:HJG22BGXX:1:11101:1063:7774 2:N:0:TAGGCATG

+

(No sequence and no quality)

[GenoMax]

Is it possible that your R1/R2 files have got out of sync? Just deleting those lines may not be enough. You should use repair.sh from BBMap to check and repair the files.

[Brian Bushnell]

You can get rid of that with BBMap's reformat program:

reformat.sh in=file.R#.fastq.gz out=reformatted.fq.gz minlen=1

Alternately, you can output directly as sam (or if samtools is installed, as bam):

reformat.sh in=file.R#.fastq.gz out=reformatted.bam minlen=1

The "minlen=1" flag will throw away 0-length reads.

[incognitos]

OK, but what about the corresponding read1?
For example if line 2 of file.R2 has read length = 0, but line 2 of file.R1 has read length >0 ?

[Brian Bushnell]

Reformat always keeps pairs together, to maintain consistency. So, both of them will be discarded.

[incognitos]

Thank you so much Brian! This seems to have solved the problem.