Glossary - Microbiome marker gene analysis with QIIME 2

action: A general term for a method, a visualizer, or a pipeline. Actions are always defined by plugins.
annotation: Additional information that can be attached to a result, such as a text file or cryptographic signature. See note and signature for more information on the respective annotation types that have been implemented, and what they can contain.
artifact: Artifacts are rachis results that are generally considered to represent intermediate data in an analysis, meaning that an artifact is generated by rachis and intended to be consumed by rachis (rather than by a human). Artifacts can be generated either by importing data into rachis or as output from an action. When written to file, artifacts typically have the extension qza. Artifacts can be provided as input to actions or exported from rachis for use with other software.
artifact class: A type of artifact that can exist in the rachis ecosystem. Examples of artifact classes include FeatureTable[Frequency] and DistanceMatrix. Every artifact has a single, immutable artifact class which defines the data that it contains (i.e., its semantic type) and the file format that it uses for storing its data internally. The class of a given artifact can be discovered by loading it with rachis-view or calling qiime tools peek, and all actions define what artifact classes (if any) they take as input, and which artifact classes (if any) they generate as output.
breaking change: A breaking change is a change to how a program works (for example, a rachis plugin or interface) that introduces an incompatibility with earlier versions of the program. This will generally require that users make some modification to how they were using some aspect of a system. For example, if a plugin method added a new required input in version 2, that would be a breaking change with respect to version 1: calling the method without that new parameter would fail in version 2, but would have succeeded with version 1. This may also be called a backward incompatible change or an API change.
collection: Collections are groupings of one or more rachis artifacts that are used as input or generated as output from actions.
conda metapackage: A metapackage is a package with no files, only metadata. They are typically used to collect several packages together into a single package via dependencies. (source)
data provenance: In the context of rachis, data provenance refers to the information automatically recorded on the steps taken to generate a specific Result. This information ensures transparency and reproducibility of analyses, and is also helpful for providing technical support to users. Data provenance information is recorded as metadata inside of qza and qzv files. It is most commonly reviewed using rachis-view or the q2cli commands qiime tools view or qiime tools replay-provenance (the latter of which uses provenance replay).
deployment: An installation of rachis as well as zero-or-more interfaces and plugins. The collection of interfaces and plugins in a deployment can be defined by a distribution.
distribution: A collection of rachis plugins that are installed together through a single conda metapackage. These are generally grouped by a theme. For example, the QIIME 2 distribution provides a collection of plugins for analysis of microbiome amplicon data, while the MOSHPIT distribution provides a collection of plugins for analysis of microbiome shotgun metagenomics data. When a distribution is installed, that particular installation is an example of a deployment.
DRY: An acronym of Don’t Repeat Yourself, and a critical principle of software engineering and equally applicable in research data management. For more information on DRY and software engineering in general, we recommend The Pragmatic Programmer, 20th Anniversary Edition. Content on DRY is available in a free example chapter here.
epoch: An identifier referencing a release cycle of rachis. Epoch identifiers represent dates (e.g., 2024.10 references the release cycle initiated in October of 2024).
even sampling depth: In studies based on DNA/RNA sequencing data, a sequencing depth that is user-selected to perform analysis at. Normalizing to that depth, for example using rarefaction, is essential for some (but not all) analyses.
export: As used in the rachis ecosystem, exporting is the process of unpacking data from artifacts into their constituent data files for use outside of rachis. When data are exported, data provenance tracking terminates and the user is responsible for tracking their own analysis steps. Learn more here.
feature: An entity observed in a study. Presence, absence, or abundance of features are generally used to describe the composition of samples. Feature is an intentionally general term, and depending on the data type in use can refer to an amplicon sequence variant (ASV), a specific taxonomic group, a functional gene category, a class of metabolite, or something else. Along with samples, features define one of the two axes in the FeatureTable artifact class and are a central concept in the rachis ecosystem.
feature identifier: See identifier.
feature metadata: See metadata.
Galaxy: Galaxy is an open source, web-based platform for data intensive biomedical research. rachis plugins can be used through a graphical interface with Galaxy, either on locally hosted hardware (see here) or on a public Galaxy server (see https://cancer.usegalaxy.org).
identifier: As used in the rachis ecosystem, identifiers are unique labels for samples and features. Chase et al. (2016) discusses considerations when defining identifiers, and more rachis-specific recommendations can be found in the metadata file format documentation.
import: As used in the rachis ecosystem, importing is the process of loading data into artifacts, at which point data provenance tracking begins and the data is assigned an artifact class. Learn more here.
interface: The layer of rachis that users (either humans or other computer software) interact with. q2cli and the Python 3 API are the two interfaces covered in Using rachis, and tutorials such as gut-to-soil and Moving Pictures additionally document use through Galaxy.
library plugin: See stand-alone plugin.
metadata: Study-specific information, generally focused on either samples or features. Metadata about samples is referred to as sample metadata and metadata about features is referred to as feature metadata, as discussed in more detail here. The shared file format for sample and feature metadata is described here. Artifacts of many classes can be viewed as metadata, a powerful approach that broadly expands the types of analyses that are available in rachis. Metadata is indexed on identifiers, and within a study identifiers must be unique (i.e., representing one entity).
method: A type of action that takes one or more artifacts or parameters as input, and produces one or more artifacts as output. For example, the filter-features action in the q2-feature-table plugin is a method.
MOSHPIT: A distribution of microbiome metagenome analysis related rachis plugins. This has previously been referred to as the metagenome distribution and the shotgun distribution. To learn more, see https://moshpit.qiime2.org.
note: A type of annotation that can be added to a result. This type of annotation can accept inline text or a path to a file whose contents will be written to the associated result, with a limit of 10 MiB.
parameter: A value that alters the behavior of an action. These are generally primitive types, such as integer or string values, as opposed to artifacts used as inputs to actions.
pipeline: A type of action that typically combines two or more other actions. A pipeline takes one or more artifacts or parameters as input, and produces one or more results (artifacts and/or visualizations) as output. For example, the kmer-diversity action in the q2-boots plugin is a pipeline.
plugin: A plugin provides analysis functionality in the form of actions. All plugins can be accessed through all interfaces. Plugins can be developed and distributed by anyone. A collection of plugins that are installed together is referred to as a distribution, and examples of distributions include QIIME 2, MOSHPIT, and tiny. Plugins that are not included in existing distributions can be discovered on rachis-library, and as a result these are typically referred to as library plugins.
primitive type: A type assigned to a parameter to an action. These are generally simple input values, such as an integer, a string, or a boolean value, as opposed to more complex data passed as input to actions through artifacts.
provenance replay: rachis functionality that enables generation of new executable code from a Result's embedded data provenance. Provenance replay can be helpful for generating detailed bioinformatics methods text, or for adapting commands used to generate one or more Results to a new data set. Learn more in Keefe et al. (2023).
Python 3 API: rachis’s Application Programmer Interface. This allows advanced users to access all rachis analytic functionality directly in Python. This can be very convenient for developing tools that use rachis as a component, or for performing data analysis without writing intermediary data artifacts to disk unless you specifically want to.
q2cli: q2cli is the original (and still primary, as of March 2024) command line interface for rachis.
QIIME 2: A distribution of microbiome amplicon analysis related rachis plugins. As this was the first rachis distribution, the term was overloaded as discussed here. This has also previously been referred to as the amplicon distribution and the core distribution. To learn more, see https://amplicon-docs.qiime2.org.
QIIME 2 Framework: See rachis.
QIIME 2 Library: See rachis-library.
QIIME 2 View: See rachis-view.
qza: An acronym for QIIME Zipped Artifact. See artifact.
qzv: An acronym for QIIME Zipped Visualization. See visualization.
rachis: The software framework that QIIME 2, MOSHPIT, and all plugins are built on. rachis provides general purpose functionality for biological data science tools, like support for different interfaces and data provenance tracking. Generally speaking, users do not need to know that they are interacting with rachis, but if your use of tools that are built on it expands, having some basic knowledge of rachis functionality will allow you to more quickly learn new tools. rachis was formerly referred to as the QIIME 2 Framework. The name rachis is preferentially stylized in monospace font: rachis.
rachis-library: A website for discovering diverse resources for the rachis ecosystem, including plugins, data resources, documentation, and workshop information. This site was formerly referred to as the QIIME 2 Library (as of this writing on 24 August 2026, rebranding is still in progress). rachis-library can be found at https://library.rachis.org.
rachis-view: A website for viewing artifacts, visualizations, data provenance, and other information from rachis results. This was formerly referred to as QIIME 2 View. rachis-view can be found at https://view.rachis.org.
rarefaction: In -omics studies based on DNA/RNA sequencing data, the process of iteratively resampling to a user-specified even sampling depth. This differs from the practice of rarefying, in that rarefying represents a single iteration of rarefaction. You can learn more in Raspet et al. (2024).
result: A general term for an artifact or a visualization.
sample: An individual unit of study in an analysis. Along with features, samples define one of the two axes in the FeatureTable artifact class and are a central concept in the rachis ecosystem.
sample identifier: See identifier.
sample metadata: See metadata.
sequencing depth: In -omics studies based on DNA/RNA sequencing data, this generally refers to the number of sequences obtained for a given sample.
signature: A type of annotation that can be added to a result. This annotation contains a cryptographic signature, providing support to ‘self-sign’ results. This allows for identity confirmation for who created a particular result.
stand-alone plugin: A plugin that is not installed by default in any distribution, but rather is installed independently. These often include cutting edge functionality. These are generally discovered through rachis-library, and are therefore sometimes referred to as library plugins, but not all stand-alone plugins are distributed on rachis-library. You may also hear these referred to as third-party plugins.
third-party plugin: See stand-alone plugin.
tl;dr: “Too long; didn’t read.” In other words, a quick summary of the content that follows.
visualization: Visualizations are rachis results that represent terminal output in an analysis, meaning that they are generated by rachis and intended to be consumed by a human (as opposed to being consumed by rachis or other software). Visualizations can only be generated by visualizers or pipelines. When written to file, visualizations typically have the extension qzv. See here for information on how to view Visualizations.
visualizer: A type of action that takes one or more artifacts or parameters as input, and produces exactly one visualization as output. For example, the scatterplot-2d action in the q2-vizard plugin is a visualizer.