Extract 1M reads from paired end fastqs

Question

[acorella]

Hi,

I have paired end reads in 2 separate fastq files. I want to take a subset of these reads for a bowtie run to get insert size. I am familiar with how to break up an individual file into 1 million reads (i.e. here: https://www.biostars.org/p/66864/)

My Question: Do I need to ensure my reads are in the same order in each file before I do this? If so, how do I do this?

Thanks!

[GenoMax]

However, do I need to ensure my reads are in the same order in each file before I do this? If so, how do I do this?

If you have not done anything to the files (other than using a paired-end aware trimming program) then the reads should be in order in R1/R2 files.

The files can be repaired as follows, if you suspect that the pairing is broken. repair.sh is from BBMap suite.

repair.sh in1=r1.fq.gz in2=r2.fq.gz out1=fixed1.fq.gz out2=fixed2.fq.gz outsingle=singletons.fq.gz

[acorella]

Thank you! That was indeed the question I was trying to ask!

Is there a quick way you can tell if the pairing is broken?

[Brian Bushnell]

reformat.sh from the same package has an option to to that:

reformat.sh in1=r1.fq in2=r2.fq vpair

That will just verify that the names indicate the reads are in the same order in each file. Incidentally, you can also randomly sample 1M pairs from them, like this:

reformat.sh in1=r1.fq in2=r2.fq out1=sampled1.fq out2=sampled2.fq samplereadstarget=1m

If your reads are overlapping, you can discover the insert size with BBMerge; if not, you'll need to use mapping.

[GouthamAtla]

duplicate of Selecting Random Pairs From Fastq? ?

[GenoMax]

And other similar threads (a subset)

How to randamly extract reads from a FASTQ file?
How To Randomly Select 20M Reads From A 200M Fastq File
Select sequences from fastq.gz file

Answer 1

[Biomonika (Noolean)]

seqtk sample with fixed seed should work for you. Take a look here:

Selecting Random Pairs From Fastq?