Running the dada2 denoise command is taking a really long time. Is it possible to run this process in parallel? In the announcement for qiime2-2017.2, it says a new feature is “Parallel processing support and other performance enhancements for DADA2”.
Thanks.
Hi @linneakh! You can pass the --p-n-threads parameter to either denoise-single or denoise-paired to control the number of threads used for processing. The parameter takes an integer (the number of threads dada2 should use when running), or, if you specify 0, it will use all available cores.
For future reference, you can pass the --help flag to any method or visualization in a plugin to get detailed information about the relevant inputs, outputs, and parameters.
$ qiime dada2 denoise-single --help
Usage: qiime dada2 denoise-single [OPTIONS]
This method denoises single-end sequences, dereplicates them, and filters
chimeras.
Options:
--i-demultiplexed-seqs PATH Artifact:
SampleData[PairedEndSequencesWithQuality |
SequencesWithQuality] [required]
The
single-end demultiplexed sequences to be
denoised.
--p-trunc-len INTEGER [required]
Position at which sequences
should be truncated due to decrease in
quality. This truncates the 3' end of the of
the input sequences, which will be the bases
that were sequenced in the last cycles.
Reads that are shorter than this value will
be discarded.
--p-trim-left INTEGER [default: 0]
Position at which sequences
should be trimmed due to low quality. This
trims the 5' end of the of the input
sequences, which will be the bases that were
sequenced in the first cycles.
--p-max-ee FLOAT [default: 2.0]
Reads with number of expected
errors higher than this value will be
discarded.
--p-trunc-q INTEGER [default: 2]
Reads are truncated at the
first instance of a quality score less than
or equal to this value. If the resulting
read is then shorter than `trunc_len`, it is
discarded.
--p-n-threads INTEGER [default: 1]
The number of threads to use
for multithreaded processing. If 0 is
provided, all available cores will be used.
--p-n-reads-learn INTEGER [default: 1000000]
The number of reads to
use when training the error model. Smaller
numbers will result in a shorter run time
but a less reliable error model.
--p-hashed-feature-ids / --p-no-hashed-feature-ids
[default: True]
If true, the feature ids in
the resulting table will be presented as
hashes of the sequences defining each
feature. The hash will always be the same
for the same sequence so this allows feature
tables to be merged across runs of this
method. You should only merge tables if the
exact same parameters are used for each run.
--o-table PATH Artifact: FeatureTable[Frequency] [required
if not passing --output-dir]
The resulting
feature table.
--o-representative-sequences PATH
Artifact: FeatureData[Sequence] [required
if not passing --output-dir]
The resulting
feature sequences. Each feature in the
feature table will be represented by exactly
one sequence.
--output-dir DIRECTORY Output unspecified results to a directory
--cmd-config PATH Use config file for command options
--verbose Display verbose output to stdout and/or
stderr during execution of this action.
[default: False]
--help Show this message and exit.
The same help text is also available on the doc site.