I noticed some reads not assigned in my dataset are well aligned to subjects when I inspect reads but these subjects have a ? for organism name.
I use the last version of MEGAN and the last mapping file : prot_acc2tax-Jun2018X1.abin . But I noticed the same problem with the previous mapping file and previous MEGAN version.
When I manually search ids in NCBI protein database, I found an organism and a classification.
I attached a screenshot with exemple
Is a bug or the mapping file is incomplete ?