deepnano_basecall1
Basecaller for oxford nanopore minion reads based on deep recurrent neural networks
Usage
usage: deepnano_basecall
[ARGS]
[reads
[reads ...]]
Overview
positional arguments:
-
reads
optional arguments:
-
-h, --help -
show this help message and exit
-
--template_net TEMPLATE_NET -
path to network which basecalls template (has reasonable default)
-
--complement_net COMPLEMENT_NET -
path to network which basecalls complement (has reasonable default)
-
--big_net BIG_NET -
path to network which basecalls 2D (has reasonable default)
-
--timing -
if set, display timing information for each read
-
--type TYPE -
One of: template, complement, 2d, all, use comma to separate multiple options, eg.: template,complement type of basecalling output (defaults to all)
-
--output OUTPUT -
output filename
-
--output_orig -
if set, outputs also Metrichor basecalls
-
--directory DIRECTORY -
Directory where read files are stored
Examples
- For SQK-MAP-006 chemistry just use:
-
OMP_NUM_THREADS=1 deepnano_basecall <list of fast5 files>
or
OMP_NUM_THREADS=1 deepnano_basecall --directory <directory with reads>
It outputs basecalls for template, complement and 2D into file named output.fasta.
- For SQK-MAP-005 chemistry use:
-
OMP_NUM_THREADS=1 deepnano_basecall --template_net nets_data/map5temp.npz --complement_net nets_data/map5comp.npz --big_net nets_data/map5-2d.npz <list of fast5 files>
Releases
Unchanged in 2 releases since Debian 10 (Buster).