I have a question about the assignment of reads to taxa. If I inspect the reads, I see that there are some question marks for reads with the highest score. How can I properly assign all of these? I’ve been using the megan-nucl-Jan201.db mapping file (I believe that’s the most recent?).
This also results in a lot of unassigned reads, while the reads still meet the given parameters (e.g. 90% identity), but it seems that the references just don’t get classified.
Do you have some idea when you will release the new mapping database? I’m asking it because when I ran my shotgun samples against a new version of NCBI-nr (november 2021) using DIAMOND and after meganized them, I realized that all eukaryota assignments in taxonomic tree were generic while the Bacteria assignments were more specific. For example, I had 32,969 assignments to Opisthokonta (in the tree) and no subclades; however, when you check this clade in the “Inspect” tool, there are several specific assignments for different species (e.g Alternaria alternata and others). Researching in the forum, I guess that a new database could solve this specific question. There is another topic in the forum that is related to this - MEGAN sorting all alignments as unclassified -