usage of output_dir is not as intended

[gregcaporaso]

Hello, I noticed this comment in your protocol:

# Files don't actually get put in output-dir

If you specify the --output-dir option, and outputs whose paths are not specified with --o- options will be put in the specified output directory with a default name. So, think of the --output-dir as the default location where outputs will be written, which you can then override for none, some, or all of the outputs by providing their `--o- options.

Some variants on your command and their results are as follows:

# write all three outputs to the asvs directory with a default file names
qiime dada2 denoise-paired \
    --i-demultiplexed-seqs paired-end-demux-trimmed.qza \
    --p-n-threads 0 \
    --p-trunc-q 2 \
    --p-trunc-len-f 219 \
    --p-trunc-len-r 194 \
    --p-max-ee-f 2 \
    --p-max-ee-r 4 \
    --p-n-reads-learn 1000000 \
    --p-chimera-method pooled \
    --output-dir ./asvs

# write 2 of 3 outputs to the asvs directory with a default file name, and put the other in a specified location
qiime dada2 denoise-paired \
    --i-demultiplexed-seqs paired-end-demux-trimmed.qza \
    --p-n-threads 0 \
    --p-trunc-q 2 \
    --p-trunc-len-f 219 \
    --p-trunc-len-r 194 \
    --p-max-ee-f 2 \
    --p-max-ee-r 4 \
    --p-n-reads-learn 1000000 \
    --p-chimera-method pooled \
    --output-dir ./asvs/ \
    --o-table some-other-directory/table-dada2.qza 

# write all three outputs to specified locations 
qiime dada2 denoise-paired \
    --i-demultiplexed-seqs paired-end-demux-trimmed.qza \
    --p-n-threads 0 \
    --p-trunc-q 2 \
    --p-trunc-len-f 219 \
    --p-trunc-len-r 194 \
    --p-max-ee-f 2 \
    --p-max-ee-r 4 \
    --p-n-reads-learn 1000000 \
    --p-chimera-method pooled 
    --o-table table-dada2.qza \
    --o-representative-sequences rep-seqs-dada2.qza \
    --o-denoising-stats stats-dada2.qza

[rhlampe]

Great! Thank you for looking this over and clarifying this!

[rhlampe]

Fixed with commit 70286c4e78a0e6e2b5302c6092ea6ee776f01db5