seqcounter 1.0.0

seqcounter becomes an installable, tested package. Everything the first version did, it still does, with the same flags and the same refseqcounts.tsv; what changed is around that.

conda create -n seqcounter -c conda-forge -c bioconda bbmap python">=3.10"
conda activate seqcounter
pip install seqcounter
seqcounter -ref refs.fasta -i samples/ -o results/

python seqcounter.py still works from a clone, with the same arguments, so existing scripts keep working.

Packaging and interface

• A seqcounter command and python -m seqcounter, installable from PyPI (and bioconda shortly).
• No Python dependency at all: pandas and psutil are gone. Only BBTools and Java are needed, as before.
• The counts table is printed to standard output, with the messages on standard error, so seqcounter ... > counts.tsv works.
• Two new reports beside refseqcounts.tsv: refseqcounts_samples.tsv (reads, bases, matched reads and the error per sample) and refseqcounts.json (parameters, software versions, references and counts).
• New options: -s/--sample, --memory, --verbose, and -j/--jobs as another name for -p/--parallel. -v still prints the version.

Counts that can be trusted

• A single fastq file is never read as interleaved: Seal guesses interleaving from the read names, and where it guessed wrong the counts were halved.
• The reference file is checked before anything runs: a sequence shorter than -k (which Seal cannot find at all, and which used to be reported as a zero count), an empty sequence, two sequences whose first word is the same (their counts were added together), a file that is not fasta.
• Sample name collisions are reported instead of being resolved silently: two files reducing to the same name used to be mixed into one entry and passed to Seal as a bogus pair.
• File names Seal cannot be given safely (spaces, quotes, $, ;, &, parentheses, commas, or xmx/xms) are linked under a neutral name; such a name could previously make the run fail, or run a command.
• Formats are recognized from the content, not only the extension, and the quality encoding is forced (qin=33).
• Seal's memory is bounded: the Java heap is sized from the reference sequences and capped at 85% of the machine divided by the parallel jobs. Before, each job reserved most of the free memory, so -p 2 could put the machine into swap.
• One sample failing no longer stops the run: its column is left empty, its error is reported, and the exit code is 1.
• The column order is deterministic, where it used to depend on which sample finished first.

Development

• 105 tests, 100% coverage, end to end on synthetic references, genomes and reads with known counts.
• GitHub Actions: tests on Python 3.10, 3.12 and 3.14 with ruff and Codecov, PyPI publishing on release, and the wiki published from wiki/.
• A wiki, a bioconda recipe, a CITATION.cff, and examples/run_example.sh, which generates its own data and counts it.

Full details in the changelog.