I have 4 paired-end files from two adjacent individuals of the same species. They are expected to be clones, and both libraries have relatively low sequencing depth:

seq1_1.fastq.gz
seq1_2.fastq.gz
seq2_1.fastq.gz
seq2_2.fastq.gz

Because coverage is low, I’m considering pooling them. Option 1: Treat them as the same biological sample using the

run

column (merge before assembly). For example:

sample,run,group,short_reads_1,short_reads_2,short_reads_platform
sampleX,1,0,seq1_1.fastq.gz,seq1_2.fastq.gz,ILLUMINA
sampleX,2,0,seq2_1.fastq.gz,seq2_2.fastq.gz,ILLUMINA

Option 2: Keep them as separate samples but assign the same group and run with

--coassemble_group

, e.g.:

sample,group,short_reads_1,short_reads_2,short_reads_platform
sample1,0,seq1_1.fastq.gz,seq1_2.fastq.gz,ILLUMINA
sample2,0,seq2_1.fastq.gz,seq2_2.fastq.gz,ILLUMINA

Then use:

--coassemble_group

Biologically, these are adjacent individuals from the same population and expected to be clones. My main goal is improving assembly quality due to low depth. Would you recommend merging them as technical runs (Option 1), or keeping them separate and co-assembling (Option 2)? Does pooling make sense biologically in this case?

(i.e., if the coverage of the symbiont is at 5x in sample 1, and 10x in sample2, you would expect all the contigs from the symbiont to follow that pattern, which you can use to fish them out)

Hi @Jim Downie or @Daniel Straub, would you mind if I ask a follow-up question? I am pooling all samples for a run, and it includes multiple individuals from multiple species, hope to resolve more symbiont mags. Would you recommend me still go with option 2 (different sample name, same group,

--coassemble_group

)?