SemiBin2
Starting with version 1.5 (officially SemiBin2 beta), installing the SemiBin package installs two scripts: SemiBin
and SemiBin2
.
They have the same functionality, but slightly different interfaces.
As of version 2.0, the older SemiBin
command is not recommended (except for backwards compability) and newer projects should use SemiBin2
.
In version 2.1, we will deprecate the SemiBin
command and introduce a more explicit SemiBin1
subcommand for backwards compatibility.
In version 2.2, SemiBin
not be installed and SemiBin1
will be deprecated.
Upgrading to SemiBin2
- If you are using the
easy_*
workflows, then they will probably continue to work exactly the same (except that you will get better results faster). - Outputs are now always in a directory called
output_bins
(unless you explicitly ask for the pre-reclustered bins to be written out with the--write-pre-reclustering-bins
option). - By default, bins are in file named as
SemiBin_{label}.fa.gz
(and compressed with gzip as the name indicates; you can change the compression with the--compression
flag).
Points 2
and 3
may require some minor modifications to wrapper scripts.
Longer list of differences between SemiBin2 and SemiBin1
The biggest different is that the default training mode is self-supervised mode.
- Output bins are now in a directory called
output_bins
(in SemiBin1, it actually depended on which parameters were used). - Output filenames are now anvi'o compatible (effectively, the default value of
--tag-output
isSemiBin
), see discussion at #123. --compression
defaults togz
(instead ofnone
)- ORF finder defaults to the
fast-naive
internal ORF finder --write-pre-reclustering-bins
isFalse
by default- To train in semi-supervised mode, you must use the
train_semi
subcommand (and there is notrain
subcommand, you must be specific:train_semi
ortrain_self
).
A few arguments that were deprecated before are completely removed:
- --recluster
: it did nothing already as reclustering is default
- --mode
: Use --train-from-many
- --training-type
: Use --semi-supervised
to use semi-supervised learning
(although that is also deprecated)