d4straub
commented
1 year ago Hi there, all question welcome!
When I look at the output files on the nexflow ampliseq github page under /qiime2/abundance_tables, you see the filtered-table.qza with a short description of "QIIME2 fragment". When I look at this filtered-table file as a qzv on qiime2 view, the only difference I can see is that the number of features and total frequency decreased from the original "table.qza" file.
I assume you are speaking about the filtered-table.qza file in the AWS results tab on the nf-core website here? Furthermore, I assume the table.qza file is from here? The difference is essentially the filters explained directly below https://nf-co.re/ampliseq/2.5.0/output#qiime2, i.e. --exclude_taxa (default is removing mitochondria & chloroplast), --min_frequency (default is off), and --min_samples (default is off).
However, it does not get rid of any samples.
Correct, the filters are removing ASVs, not samples. The only way to remove samples currently was introduced in version 2.5.0 with https://github.com/nf-core/ampliseq/pull/538, i.e. --diversity_rarefaction_depth and --ancom_sample_min_count.
Is there a specific reason you want to get rid of samples? And based on what specifics (e.g. total sample counts, sample name, ...)? A sample filtering step at other points of the pipeline based e.g. on total sample counts would be relatively easy to add, but I am not sure why it would be needed.
Sorry if this is a silly question but I wanted to know what parameters the pipeline chose to filter the "filtered-table.qza". When I look at the output files on the nexflow ampliseq github page under /qiime2/abundance_tables, you see the filtered-table.qza with a short description of "QIIME2 fragment". When I look at this filtered-table file as a qzv on qiime2 view, the only difference I can see is that the number of features and total frequency decreased from the original "table.qza" file. However, it does not get rid of any samples.