Hello,

I would like to assemble the reads that did not align to my reference genome. I understand that I can gather the unmapped reads with

samtools view -f 12 -F 256 aligned-sorted-deduplicated.bam > unmapped.bam

where -f12 = keep read unmapped, mate unmapped and -F256 = skip not primary alignment. I am essentially keeping only the pairs that did not match. Then I need to re-sort the reads with

samtools sort -n unmapped.bam unmapped-sorted.bam

The manual indicates to use fastq files, so I need to extract them with

bamToFastq -i unmapped-sorted.bam -fq R1.fastq -fq2 R2.fastq

and then assemble with

spades.py -1 R1.fastq  -2 R2.fastq  -o someFolder

I would like to ask:

1. Is this procedure correct?
2. Can I compress the fastq files directly?
3. Can I use the unmapped-sorted.bam without making the fastq files and if yes how?

Thank you