splitting fastq files output size issue

Question

[adrilu.romero]

I have a 22 Gb file that has paired end reads merged. In other words, in a single file, I have all R1 and R2s. After splitting the file to get 2 files (to input to Trinity), one in which I have all R1s and another one will all R2s, why are my single files much smaller than the original file? R1 file is 1.22 gb and R2 file is also 1.22 gb. Thanks!

[Eric Lim]

How did you split the file?

[adrilu.romero]

using:

paste - - - - < test.fq \ | tee >(awk 'BEGIN{FS="\t"; OFS="\n"} {if (match($1, " 1:N")) print $1,$2,$3,$4}' > test.r1.fq ) \ | awk 'BEGIN{FS="\t"; OFS="\n"} {if (match($1, " 2:N")) print $1,$2,$3,$4}' > test.r2.fq

[Eric Lim]

I think that's unnecessarily complicated. Have you tried some of the methods from Fastq Splitter For Paired End Reads

[Pierre Lindenbaum]

I think that's unnecessarily complicated

yup

  rm -f R1.fq R2.fq && cat R0.fq  | awk '(NR%4==1) { out = (index($0,"1:N")?"R1.fq":"R2.fq");} { print >> out;}'

are you sure ALL your reads contain 1:N or 2:N ?