I'm trying to analyse some shotgun soil metagenomics data. Running the forward read of one of my samples through DIAMOND and MEGAN results in around 70% of the reads being unassigned in each of the functional modules (i.e. SEED, EGGNOG etc.). Almost all of the reads are assigned in the taxonomic module.
Is this kind of result expected with a shotgun metagenomics sample, or is something going wrong in the Meganizing step?