Install

The recommended way to install SeqFu is from Bioconda, on Linux and macOS:

conda install -c conda-forge -c bioconda "seqfu>=1.30"

Other options (pre-compiled binaries, building from source) are described in Installation. Check that everything works with:

seqfu version

First commands

SeqFu reads plain or gzipped FASTA and FASTQ files, and most commands behave like the Unix tools you already know, but are aware of sequence records:

# How many reads? Paired files are detected and checked together
seqfu count reads_R1.fastq.gz reads_R2.fastq.gz

# Length statistics (N50, auN, min/max...) as a terminal table
seqfu stats -n assembly.fasta

# The first 5 records, or one every 100 reads
seqfu head -n 5 reads_R1.fastq.gz
seqfu head -n 5 --skip 100 reads_R1.fastq.gz

# Records containing a (degenerate) primer, on either strand
seqfu grep -o CCTACGGGNGGCWGCAG amplicons.fasta

Typing seqfu alone prints the list of subcommands, and every subcommand has its own help, for example seqfu stats --help. seqfu cite prints the paper to cite.

Core tools and utilities

SeqFu ships two kinds of programs:

• Core tools are subcommands of the main binary: seqfu stats, seqfu grep, seqfu interleave… Their command-line interface is covered by an extensive test suite, so options are stable across releases.
• Utilities are standalone programs, usually with an fu- prefix (fu-cov, fu-primers…). They cover more specialised tasks. A few of them are now compatibility wrappers for commands that moved into the core (fu-orf runs seqfu orf, for instance).

The Tools catalogue lists both, and can be filtered by task, input type and kind.

Where next

• Overview of the commands: a guided tour, grouped by task
• Common conventions: standard input, compressed files, paired-end naming
• Recipes: short, end-to-end workflows