huge false positive ratio

[farhadm1990]

Hi guys,

I am running Kaiju on paired-end fastq file for whole genome shotgun metagenomics mock community. I have preprocessed the reads and am using refseq_ref database prebuilt index. I run it as follows for instance:

kaiju -t /mnt/storage/farhad/database/kaiju_db/refseq_ref/nodes.dmp
-f /mnt/storage/farhad/database/kaiju_db/refseq_ref/kaiju_db_*.fmi
-i out_metphl/qc/krak_no_host/D0165Nr131_NoHost_R1.fastq.gz
-j out_metphl/qc/krak_no_host/D0165Nr131_NoHost_R2.fastq.gz
-s 80 -e 2 -z 4 -o kaiju_class/mock_kaiju.out

kaiju2table -t /mnt/storage/farhad/database/kaiju_db/refseq_ref/nodes.dmp -n /mnt/storage/farhad/database/kaiju_db/refseq_ref/names.dmp -r species -p -o kaiju_count.tsv mock_kaiju.out

kaiju_count.tsv contains >3000 species instead of only 14 species.

I have already classified them by Metaphlan 4 and it returned the same hit with the same rel_abund, so the mock sample is clean contains only mock species.
I was wondering if I am doing something wrong here?

Appreciated in advance.

Farhad