Are you an LLM? You can read better optimized documentation at /micos-2024/en.md for this page in Markdown format
Open-source metagenomics whitepaper
MICOS-2024
A documentation experience rebuilt as a technical brief: not a checklist of commands, but a guided map of how this repository turns raw sequencing input into reproducible, reviewable microbiome analysis outputs.
The site is optimized for demanding readers: interviewers, maintainers, and senior open-source engineers evaluating whether the project is coherent beyond its README.
Repository truth, execution boundaries, architecture intent, and research lineage.
What is deliberately surfaced
The current stable CLI core, the broader workflow assets, and the gap between ambition and implemented runtime surface.
How to read it
Start in Academy, then move into Architecture, then drop into Guides or Research depending on whether you are operating or auditing.
Primary runtime
CLI-first orchestration
The stable surface is a Click CLI backed by Python modules, with shell wrappers kept as compatibility layers.
Workflow posture
WDL + containers
The repository carries step-level WDL assets, Singularity definitions, and a Docker Compose example for reproducible environments.
Reader outcome
Interview-grade clarity
This site is organized to let a reviewer understand scientific scope, software boundaries, and operational trade-offs quickly.
Pipeline narrative
The platform is easier to trust when the execution chain is visible.
MICOS-2024 is best understood as an analysis story with four checkpoints: input hygiene, taxonomic evidence, ecological interpretation, and report-facing outputs.
Vue SVG component for zero-delay theme switching, single-source maintenance.
Reader signal The pipeline is presented as a sequence of evidence transformations, not just a list of tools.
Engineering signal The repository carries both stable CLI entrypoints and broader workflow assets. This site distinguishes them instead of flattening everything into one surface.
Operational signal Several advanced analyses live under scripts/ as specialist tools. They matter, but they are not described as part of the same stability contract as the CLI core.
Stage 01
Quality control and host depletion
FastQC and KneadData frame the entry gate, producing cleaner reads before downstream interpretation begins.
Functional profiling and summarization consolidate pathways, annotations, and report-facing deliverables.
Functional tables
Auxiliary scripts
HTML-oriented summaries
System anatomy
Repository layers, not just pages
A reviewer should be able to map the docs to the codebase: entry commands, Python modules, workflow definitions, configuration templates, containers, and validation surfaces.
The project operates across a CLI layer, Python orchestration modules, workflow assets, and power-user scripts.
Stable core
micos/cli.py exposes full-run, validate-config, and module-level commands for quality control, taxonomy, diversity, functional annotation, and summarization.
Workflow assets
steps/, deploy/, and containers/ extend the platform into reproducible environments and step-level orchestration patterns.
Research posture
The project benefits from established microbiome tooling, then attempts to wrap it into a more coherent end-to-end suite. That lineage is explicit in the Research section.
Execution chain
A quick mental model
The runtime stack spans more than one abstraction level. This diagram makes the split explicit so contributors and reviewers can reason about where each concern lives.
Research grounding
The project inherits credibility from the ecosystem it assembles.
MICOS-2024 is not a blank-slate invention. It is an integration effort sitting on top of well-cited microbiome tooling. That is a strength, and this site treats it as one.
