Does the 'amptk filter' remove non-representative OTUs ? about amptk

[nextgenusfs]

Did some samples get dropped? I think it looks at the raw counts prior to any manipulation of the OTU table.

[bsmoda]

No samples get dropped, but the input of 'amptk filter' are the otu table and the otu fasta file generated from 'amptk cluster'. I thought that the filter command would drop otus (from otu_table) which didn't pass the number of --min_reads_otu parameter.

[nextgenusfs]

The filter is done by summing each row (OTU) and dropping those that are less than the --min_reads_otu parameter. This is done prior to any other filtering, i.e. if a mock community is used for filtering these calculations are done first. Do you have a small test case you can show me that isn't working?