Sadly I don't think your script can solve my problem. What I am interested in is something like this:
But instead of a taxonomic annotation to the corresponding Feature ID, I need taxonomic annotation directly to the corresponding Read ID or at least the whole original read existing in the sample fastq file.