Can you explain pooling? Paired-end reads consist of R1 and R2 and these must be mapped with a dedicated paired-end aligner. You cannot simply combine R1 and R2 and treat them as single-end reads.

If you are talking about interleaving paired-end reads from individual files into a single file, BBMap is your friend:

reformat.sh in=reads_1.fastq in2=reads_2.fastq -out=paired_reads.fastq