v2.3.0 - 2026-09-23

Added

• #95 - Accept gzipped input files throughout the pipeline, for every samplesheet column and individual file parameter alike (#76) (by @erikrikarddaniel)

Changed

• #95 - Check the file name extension of every input file parameter, in samplesheets and individual parameters alike, instead of accepting any name; download URLs with no file name extension at all are still accepted (by @erikrikarddaniel)
• #95 - Write the Clustal Omega and MAFFT alignments gzipped, like the rest of the pipeline's intermediate files (by @erikrikarddaniel)

Fixed

• #98 - Search with the profile named in extract_hmm also when a row's target equals the name of its hmm file, instead of silently searching with every profile in the file (by @erikrikarddaniel)
• #97 - Correct the usage and output documentation (wrong parameter names in examples, wrong output file names and descriptions), document the extract_hmm and min_bitscore sample sheet columns and the search outputs, and cite Clustal Omega and DuckDB in the MultiQC methods description (by @erikrikarddaniel)
• #96 - Run MAFFT with --keeplength as documented, so that query sequences with insertions no longer widen the reference alignment (by @erikrikarddaniel)
• #94 - Memory usage in hmmer/hmmrank, by rewriting it from an in-memory R script to streaming DuckDB SQL (#93) (by @erikrikarddaniel)

Dependencies

software	previously	now
duckdb-cli		1.5.5
gawk		5.3.1
gzip		1.13
nft-fasta		1.0.0
nft-utils	0.0.3	1.2.0
r-base	4.2.3
r-data.table	1.14.8
r-dtplyr	1.3.1
r-stringi	1.8.4
r-stringr	1.5.1
r-tidyverse	2.0.0

Thanks

Thanks to everyone who reviewed PRs along the way to this release: @piplus2, @pinin4fjords, @Joon-Klaps, @sguizard and @Aaramis.