Reads directly assigned to taxon level above the level specified in Bracken kreport

[agatadziegiel]

Hello,

I used Bracken (v2.2) to re-estimate reads at species level. In the resulting kreports, most of the abundances in the third column (reads directly assigned to taxon level) are 0 for anything above species, as expected. However, there are some taxa with non-zero values including G, G1, F1 etc. I am confused as to why these are being retained, particularly those in non-intermediate ranks (e.g. G).
I also ran Bracken with a read threshold of 10, but I can see some abundances of less than 10 in the kreport. These do not appear in the breport, however as I am combining the kreports with kraken-biom for downstream analysis, I would like to know why this may be happening? I would highly appreciate any help with this.

Thank you in advance!