Dear all,

I am dealing with amplicon sequencing from an environmental sample to find genotype of Nororvirus.

I have obtained contigs with SSAKE assembling de novo reads already selected via Bowtie2 alignment against reference (VP1 region of genotypes).

Looking at a sample of results below, my question is which coverage threshold is good to consider a contig reliable?

6x with a 98% alignment is a reliable contig to claim a specific virus genotype?

query_contig         nr of reads (mean_coverage)    subject_var %identity
----------
2017_GI_contig4    ----     545(461x)                    ---      I.1|M87661    --- 90.941
----------
2017_GI_contig18  ---     197(171x)              ---          I.2|L07418     ----97.674
----------
2018_GI_contig10  ---     75(61x)                    ---      I.1|M87661    --- 91.030 
----------
2018_GI_contig48    ---     7(6x)                ---                  I.2|L07418    --- 98.007
----------
2018_GI_contig3    ---     380(296x)     ---                      I.4|AJ277616  --- 95.652