How to split forward and reverse reads from one fastq file?

Question

[phongphak.06]

Hi, I have a metagenomic reads simulated by Grinder like this (example):

@1/1 reference=NZ_CP009501.1 position=complement(1484183..1484382) description="Methanosarcina thermophila TM-1, complete genome"
TAGTCAGTTCTGGCACGAATCCTGTCACACTCTACTTATTTTAAAATCCTTATTTATCACGATTTATATTTATAATTTATTATCAGTAAGATTATATTTATATACTTATGGTGTTGTATTATTCCCTGCATAAGTAATCCGATCCTATCAGTTTTCAGGCAGGACTTTGATATTTAGGGGGTCTAAAGTCCGAAAACCAG
+
????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????
@1/2 reference=NZ_CP009501.1 position=1484083..1484282 description="Methanosarcina thermophila TM-1, complete genome"
CTGTAGAAATACATTGAGAAATATCTTCTCTTATGTTTTCTTGCGTTTTTTATTTTCAGTAATCACTTACCTGAATGAGATACAACAATTCAATGATCATCTGGTTTTCGGACTTTAGACCCCCTAAATATCAAAGTCCTGCCTGAAAACTGATAGGATCGGATTACTTATGCAGGGAATAATACAACACCATAAGTATA
+
????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????

So I think /1 and /2 may represent forward and reverse reads

Answer 1

[Devon Ryan]

awk '{if(NR%8  >= 1 && NR%8 < 5) print}' file.fastq > file_R1.fastq
awk '{if(NR%8 == 0 || NR%8 >= 5) print}' file.fastq > file_R2.fastq

[phongphak.06]

Thank you! It's a nice idea, but it's not working correctly.

Anyway, I had tried this awk '{if(NR%8 < 4) print}' file.fastq > file_R1.fastq and it's worked, while I have no idea how to correct the another, since > 4 cannot extract the quality of reads (> 3 and > 5 are still not working)

[Devon Ryan]

Ah, I've updated my reply, I made an off-by-one error.

Answer 2

[Brian Bushnell]

Alternatively, with BBMap:

reformat.sh in=file.fastq out=r#.fastq