Installer kallisto avec Homebrew, apt, Nix

Liens du projet

• Page d'accueil
• Dépôt

Commandes et alias

• kallisto

Historique du projet

The public GitHub repository was created on 2014-11-13, and GitHub releases show v0.42 labeled as the first release on 2015-05-06. The 2016 Nature Biotechnology paper by Bray, Pimentel, Melsted, and Pachter introduced kallisto as a near-optimal probabilistic RNA-seq quantification program and described it as much faster than earlier approaches.

The README and manual place the tool in a reproducible scientific workflow: build an index from a transcriptome, run quantification on paired-end or single-end reads, optionally bootstrap estimates, and analyze bulk RNA-seq output with sleuth. The repository also points to scripts for reproducing the paper results, reinforcing its role as both software and a published method.

Later development extended kallisto's neighborhood beyond bulk RNA-seq. The README links kallisto with bustools and the kallisto-bustools ecosystem for preprocessing single-cell RNA-seq data, and releases in the 0.50.x and 0.51.x era reference long-read kallisto work before v0.52.0 restored genomebam and pseudobam functionality in 2026.

Historique d'adoption

kallisto became widely packaged because it occupies a high-value scientific niche: a single executable that made transcript quantification fast enough for routine laptop and cluster workflows. The batch metadata lists Homebrew, Debian, Ubuntu, and Nix packages, while upstream releases provide binaries and source for research environments that pin methods by version.

Its adoption is also tied to downstream method ecosystems. sleuth gave users a statistical analysis path for kallisto-quantified bulk RNA-seq, while bustools and kallisto-bustools made the pseudoalignment idea part of single-cell preprocessing workflows.

Modes d'utilisation

A typical kallisto workflow builds a transcriptome index and then runs `kallisto quant` on sequencing reads to produce abundance estimates. The manual documents related commands such as `index`, `quant`, `bus`, `pseudo`, `merge`, and `h5dump`, making the package useful both for direct quantification and for integration into larger pipelines.

Because kallisto is method software as much as a utility, users often cite the paper, pin the executable version, and keep index-building inputs alongside analysis scripts.

Pourquoi les passionnés de paquets s'y intéressent

kallisto is a classic example of research software crossing into general package managers: it is small enough to install like a normal CLI, but scientifically important enough that packaging, versioning, and binary reproducibility matter.

For package indexes, it also demonstrates the challenge of scientific CLI metadata: the package name is simple, but the real provenance includes a paper, a lab website, release binaries, and companion tools such as sleuth and bustools.

Chronologie

• 2014: pachterlab/kallisto repository created.
• 2015: v0.42 first release published on GitHub.
• 2016: Near-optimal probabilistic RNA-seq quantification paper published in Nature Biotechnology.
• 2024: v0.51.0 release named lr-kallisto.
• 2026: v0.52.0 release restored genomebam and pseudobam features.

Related projects

• sleuth is the companion analysis project for kallisto-quantified bulk RNA-seq output.
• bustools and kallisto-bustools extend the kallisto ecosystem into single-cell RNA-seq preprocessing.
• The kallisto paper analysis repository contains scripts for reproducing the published results.

Sources

• https://api.github.com/repos/pachterlab/kallisto
• https://github.com/pachterlab/kallisto
• https://github.com/pachterlab/kallisto/releases
• https://github.com/pachterlab/kallisto_paper_analysis
• https://github.com/pachterlab/sleuth
• https://pachterlab.github.io/kallisto/manual.html
• https://www.kallistobus.tools/
• https://www.nature.com/articles/nbt.3519