sra toolkit fastq-dump for paired end read set

[jgibbons1]

I've read the threads out there but haven't been able to solve this problem.

I am trying to separate a paired end sra formatted file to the corresponding two fastq mate pair files.

I've used both:

fastq-dump --split-files MyFile.sra
fastq-dump --split-3 MyFile.sra

The output consists of 2 files, but one of the files has a read length of 75, while the other has a read length of 9. Obviously, not working correctly.

Here's what it looks like:

@SRR514860-split_test.1 SOLEXA3_0001:5:1:1024:2576 length=75
NTCGTTACAATATCCACCCTGTCCCCGAAGAATGCTCTTGNNNAGNNNNNNNNNNNNNNNNNTNNNNNNNNCTNA
+SRR514860-split_test.1 SOLEXA3_0001:5:1:1024:2576 length=75
#))))+++(*7AAAAAA75A33303AAAA##############################################
@SRR514860-split_test.2 SOLEXA3_0001:5:1:1024:6017 length=75
NCGCGTTCCGAAGCCATTTTCATCAAAATCTCGTGAAAAANNNATNNNNNNNNNNNNNNNNNANNNNNNNNTTNG
+SRR514860-split_test.2 SOLEXA3_0001:5:1:1024:6017 length=75
#++('+****5AAA83353.3388803.35AAAA#########################################
@SRR514860-split_test.3 SOLEXA3_0001:5:1:1024:19859 length=75
NGCGCCCCCTGTGCAGAGGTACTATTGCTGCTGCTGCTGCNNNTGNNNNNNNNNNNNNNNNNANNNNNNNNACNG

==> SRR514860-split_test_2.fastq <==
@SRR514860-split_test.1 SOLEXA3_0001:5:1:1024:2576 length=9
NNNNNNNNN
+SRR514860-split_test.1 SOLEXA3_0001:5:1:1024:2576 length=9
#########
@SRR514860-split_test.2 SOLEXA3_0001:5:1:1024:6017 length=9
NNNNNNNNN
+SRR514860-split_test.2 SOLEXA3_0001:5:1:1024:6017 length=9
#########
@SRR514860-split_test.3 SOLEXA3_0001:5:1:1024:19859 length=9
NNNNNNNNN


Has anyone run into something similar before? Could this be due to the way the file was submitted and not necessarily the sra toolkit?

Thanks,

John

[aaronh]

I had a similar problem and the only way to fix it was to ask SRA to reformat it on their end. This required the input of the original submitter to verify how it should be split.

[jgibbons1]

Thanks aaronh...SRA identified a problem with the file.