Command Cheat Sheet

Command cheat sheet

The whole command line on one page: what each command is for, how to call it, and every option. Each section links to the page that explains it in depth.

<path|-> means a file, or - for standard input. Every command reads plain FASTQ and gzip alike.

The commands

count

Counts records, as fast as the file can be read. See Counting and statistics.

z-fastq count <path|-> [<path|-> ...]
z-fastq count lane1.fastq.gz lane2.fastq.gz

stats

Summarizes reads, bases, length, GC, and quality. See Counting and statistics.

z-fastq stats [--json] <path|-> [<path|-> ...]
z-fastq stats --json reads.fastq.gz > stats.json

check

Validates every record, or every pair with --paired or --interleaved. Prints nothing on success. See Validation and Paired reads.

z-fastq check [--json] <path|-> [<path|-> ...]
z-fastq check --paired R1.fastq.gz R2.fastq.gz
z-fastq check --interleaved interleaved.fastq.gz

sample

Writes a reproducible subset to standard output, by fraction or by exact count. Exact counts need regular files. See Sampling.

z-fastq sample --fraction P [--seed S] <path|->
z-fastq sample --count K [--seed S] path
z-fastq sample --paired --count 100000 --seed 7 R1.fastq.gz R2.fastq.gz > sample.fastq

interleave and deinterleave

Join R1 and R2 into one interleaved stream, or split one into two new files. See Interleaving.

z-fastq interleave R1.fastq.gz R2.fastq.gz > interleaved.fastq
z-fastq deinterleave --out1 R1.fastq --out2 R2.fastq interleaved.fastq.gz

process

Validates, trims, filters, and measures pairs into two new files, with an optional report and MultiQC files. See Processing paired reads and Quality control.

z-fastq process --out1 OUT1 --out2 OUT2 [options] R1 R2
z-fastq process --gzip --adapter1 CTGTCTCTTATACACATCT --adapter2 CTGTCTCTTATACACATCT \
  --trim-poly-g 10 --min-length 30 --report sample.json \
  --out1 trimmed_R1.fastq.gz --out2 trimmed_R2.fastq.gz R1.fastq.gz R2.fastq.gz

Options

Everywhere

• -h, --help: help, before or after a command. -V, --version: the version, before a command.
• --max-line-bytes N: the longest line accepted, 16 MiB by default (Limits).
• --: ends the options, for a path that starts with -, as in z-fastq count -- ./-odd.fastq.

An option that takes a value may be given only once; a repeat is an error rather than the last one winning.

Records and pairs

• --alphabet iupac|acgtn: which sequence letters pass, iupac by default. Every command but count and stats takes it.
• --paired, --interleaved: for check and sample, read two mate files or one interleaved file.
• --pair-names illumina|exact: how mate names are matched, illumina by default (Paired reads). For every pair command and process.
• --json: machine-readable output from stats and check (Automation).

Sampling

• --fraction P: keep each record with chance P, in one pass.
• --count K: keep exactly K records, in two passes.
• --seed S: the seed, 11 by default.

Processing

Outputs:

• --out1 PATH, --out2 PATH: the two new FASTQ files.
• --gzip, --gzip-preset fast|even|dense: compress them, with the fast preset by default.
• --report PATH: the JSON report. --qc-tables DIR and --qc-summary PATH: files for MultiQC.
• --max-pairs N: stop after N pairs.

Trims, each off unless given:

• --umi-length N --umi-mates 1|2|both, --umi-skip S: move an N-base UMI from the read starts into both names, and remove S spacer bases after it.
• --adapter1 A1 --adapter2 A2: adapters and mate overlaps.
• --trim-poly-g N: poly-G tails, scanning at least N bases.
• --trim-tail-window W --trim-tail-quality Q: low-quality 3' ends.

Filters, each off unless given:

• --low-quality Q --max-low-quality-percent P: too many bases below quality Q.
• --max-n N: more than N N bases.
• --min-length N: shorter than N bases.

Measurements:

• --contaminants PATH: name the likely source of overrepresented sequences.
• --pair-duplication: the pair duplication rate.

Processing gives each option's range, and Quality control says which options need each other.

Related pages

• Getting started: the main commands, one at a time.
• Automation: exit statuses and JSON output.
• Recipes: complete commands for everyday jobs.