This is a test version of Biostars. For the public version, visit https://www.biostars.org.
How to process paired end data (using fastq-dump) to get Fw and Rv files

Hi guys

I downloaded a fastq file using the fast-dump command (sra toolkit) to get paired end data that I want to analyse. However, the fastq file comes up as one file (was expecting two; Fw and Rv). I want to use trimmomatic which needs two input files. How do I get around this?

New to the scene as you can tell!

Thanks in advance!

fastq rna-seq sratoolkit trimmomatic
• 5,078 views • link
updated by Ram Ram • written by 2405592M 2405592M

Hi guys! Sorry this is a few months later but my follow up question is related to this thread. Is it at all possible that the above SRR fastq file is in fact an interleaved fastq file. When I more the SRR file, this was the first few lines:

@SRR1909108.1 1 length=151
TGCTCTGATGAAATCACTAATAGGAAGTGCCGTCAGAAGCGATAACTGACGAGGACTACTCCTGTCTGATTAGATCGGAAGAGCACACGTCTGAACTCCAGTCACACTTGAATCTCGTATGCCGTCTTCTGCTTGAAAAAAAAAATTTTTT
+SRR1909108.1 1 length=151
A1AAAFFFFFFFGGGGFG31FDB1110B11A0EA00111A00B//BF11A/////AF0BGEAG1GAG11DG211DB/////00>1F0B/B?G/@DFGHGE1F@1@@1BF111B1FD0?/F0BB/EGHGDG1DGG1BBDGFCC?########
@SRR1909108.2 2 length=151
TTCGTGATCGATGTGGTTACGTCTTTCTATTTCTTATTTTCACTCTTCTTTACTCCATTCTCTCTTTTTTCTCTTTTTCCTTCTTCTTCTTTTTTAAATTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTCTTCTCTTTTTT
+SRR1909108.2 2 length=151
AA@AAA1FFA?ADF11F1111B0003AD333BA3122222222211A1B112AB1A11222221BF111/0A22D1D1011D1FG2B12D@############################################################

To my follow up point, if these fastq file are indeed interleaved, what would be the best way to trim off the 5' and 3' adapter sequences? Would you have to treat the fastq file as single end reads or is there another proceedure?

• 1 views • link
written by 2405592M 2405592M

SRR1909108 is a single end sample (here is ENA entry for it) so there is no interleaving here. Those are just reads in sequence (1,2,3,4 etc).

• 1 views • link
written by GenoMax GenoMax

1 answer

You need --split-3, also use ENA rather than SRA if you can, it's much faster.

• 1 views • link
written by Devon Ryan Devon Ryan

Thanks Devon! 2 questions. 1) what would the command be? I've tried fastq-dump --split-3 SRR1909107 but I'm still getting 1 fastq file ? 2) With regards to the ENA, can I download directly from the command line or would I have to manually download these files from the ENA website? Appreciate the help!

• 1 views • link
written by 2405592M 2405592M

The person who uploaded those samples either mislabeled them or only uploaded one of the two reads, it's unclear which. I suggest you contact whoever uploaded that and ask them.

You can use wget or curl or ascp with ENA too, just like SRA. The main difference is that you will directly get fastq files and not the silly SRA files.

• 1 views • link
written by Devon Ryan Devon Ryan

SRR1909107 is indeed single-end. Not uncommon that people mislable files that are uploaded to the NCBI. Also not uncommon that some lane replicates would be paired and other single, because who cares about confounding effects and things, right :-D Anyway, for the ENA, there is a good documentation for downloads here.

• 1 views • link
written by ATpoint ATpoint

Log in to answer this question.

More posts like this