Skip to content

Choosing a FASTQ Tool

github-actions[bot] edited this page Nov 16, 2026 · 2 revisions

Choosing a FASTQ tool

No FASTQ program is right for every job, and z-fastq is not trying to be. I think about the choice in terms of what a workflow needs, not which tool wins a race.

Why z-fastq exists

Before the expensive steps of an analysis, a workflow usually needs plain answers to small questions:

  • Are these records valid, and do the two files still describe the same pairs?
  • How many reads and bases are there, and how good are they?
  • Can I take a repeatable subset without loading the file?
  • What exactly happens when a record fails halfway through?

z-fastq is built for those questions: a small, fast tool whose failures are as well defined as its results. It is narrower than a general toolkit on purpose.

Use z-fastq when

  • You need counting, statistics, validation, sampling, or pair layout changes, on plain or gzip input alike.
  • You want exact behavior for pair names, error codes, exit statuses, and output files.
  • Memory should stay small and bounded on any input.
  • You want adapter, poly-G, and quality trimming with fastp's rules, and QC that MultiQC can read, from one streaming pass.
  • You want the reader, writer, or statistics inside a Zig program.

Use another tool when

  • fastp for the full short-read preprocessing set: deduplication, base correction, pair merging, UMIs taken from the index reads, automatic adapter detection, and its own HTML report. z-fastq follows fastp's rules for the trims and filters it has, and moves a UMI from the start of the reads into their names as fastp does; it does none of the others listed here.
  • FastQC for its HTML report and interactive viewer, Kmer Content, and quality encodings other than Phred+33. z-fastq measures FastQC's modules and writes their data, not the report.
  • SeqKit for a broad FASTA and FASTQ toolbox: many subcommands, more compression formats, threads, searching, and conversion.
  • seqtk for a compact classic with conversion, subsampling, trimming, and subsequences. If a workflow already depends on its flags and output, switching may not be worth it.
  • z-fasta for FASTA: indexing, extraction, validation, and inspection of uncompressed FASTA, built around the same priorities as z-fastq.

References lists these and the other tools z-fastq is compared with, and how each comparison is made.

Related pages

Clone this wiki locally