Releases: gbouras13/dnaapler
Releases · gbouras13/dnaapler
v1.1.0
- Adds support for reorienting contigs where the gene of interest spands the contig ends #90. Thanks @marade @oschwengers.
- Specifically, this is done by rotating each contig in the input by half the genome length, then running MMseqs2 for both the original and rotated contigs. The MMseqs2 hit with the highest bitscore across the original and rotated contigs will be chosen as the top hit to rotate by, therefore enabling detection of partial hits (on the original contig) that span the contig ends.
- This has only been implemented for
dnaapler all
(this should be the command used by 99% of users).
v1.0.1
v 1.0.0
- BREAKING CHANGE -
dnaapler
now usesMMseqs2 v13.45111
rather thanBLAST
. You will need to install MMseqs2 if you upgrade (if you use conda, it should be handled for you). The CLI is identical. - There are 2 reasons for this:
- Users reported problems installing BLAST on MacOS with Apple Silicon (see e.g. here). MMseqs2 works on all platforms and is dilligently maintained.
- MMseqs2 is much much faster than BLAST (what took BLAST a few minutes takes MMseqs2 seconds). We probably should have written
dnaapler
withMMseqs2
to begin with.MMSeqs2 v13.45111
was chosen specifically to ensure interoperability with pharokka
- The alignment results may not be identicial to
dnaapler v0.8.1
(i.e. they might find different top hits), but the actual reorientation is likely to be identical (at least in my tests). Please reach out or make an issue if you notice any discrepancies
For example - on my machine (Ubuntu 20.04, Intel i9 13th gen 13900 CPU with 32 threads), for a Staphylococcus aureus genome with 1 small plasmid, dnaapler -i staph.fasta -o staph_dnaapler -t 8
took ~129 seconds wallclock with v0.8.1
using BLAST
, while it took ~3 seconds wallclock with v1.0.0
using MMseqs2
.
v0.8.1
v0.8.0
v0.7.0
v 0.6.0
- Fixes bug where if the starting gene (
dnaA/terL/repA
) was on the reverse strand, and the top BLAST hit did not find the start codon, it would reorient the genome to begin at the end of the starting gene, not the start as desired i.e. the terL would be the last gene in the output, not the first. Thanks @susiegriggo
v0.5.2
v0.5.1
- With
dnaapler all
, adds the reoriented gene to the header (thanks @ammaraziz #67 ) - Adds
--db
parameter todnaapler all
allowing specifying a subset of genes to make up the database. In particular, if you have bacteria and plasmids,--db dnaa,repa
should speed up Dnaapler's runtime quite a bit (thanks @oschwengers #63 )
v0.5.0
v0.5.0 is the JOSS Release, as published in the Journal of Open-Source Software.
Add some changes with error and logging, and other minor improvements.