[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M75FAXGHDQNE4J9AC2X1VGKK":3,"skill-same:01M75FAXGHDQNE4J9AC2X1VGKK":87},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":29,"commitSha":30,"skillDirectory":31,"forksCount":32,"repositoryHtmlUrl":33,"repositoryHomepageUrl":6,"ownerHtmlUrl":34,"ownerBlogUrl":35,"ownerDescription":6,"updatedAt":36,"install":37,"resources":50},"01M75FAXGHDQNE4J9AC2X1VGKK","alterlab-qiime2-amplicon",null,"Runs 16S\u002FITS amplicon (microbiome) analysis with the QIIME 2 amplicon distribution (2026.1; renamed to \"qiime2\" in 2026.4) in the correct order: manifest import, cutadapt trim-paired primer removal BEFORE dada2 denoise-paired (trunc-len chosen from the demux quality .qzv), feature-classifier classify-sklearn against a version-matched SILVA 138 or Greengenes2 classifier, and diversity core-metrics-phylogenetic — teaching the .qza\u002F.qzv artifact-and-provenance model and the 2026.1 feature-table summarize change (the former summarize_plus). Use when the request mentions QIIME2, QIIME 2, qiime, 16S, 18S, ITS, amplicon, microbiome, ASV, DADA2 denoising, feature table, taxonomic classification, or core-metrics diversity. For downstream alpha\u002Fbeta diversity, PCoA, and PERMANOVA on the exported feature table prefer alterlab-scikit-bio; this is conda-only (no pip install). Part of the AlterLab Academic Skills suite.","stanfish06\u002Fskillquarium","stanfish06","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F93099613?v=4","01M75F1CK4ZW109ZK3T26VN52F","Zhiyuan","01M75F1CM53VH6F79DZ8GXKKES",12,"MIT","https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium\u002Fblob\u002Fmaster\u002Fskills\u002Falterlab-qiime2-amplicon\u002FSKILL.md",[],[],"2026-11-12 13:55:43",false,"---\nname: alterlab-qiime2-amplicon\ndescription: 'Runs 16S\u002FITS amplicon (microbiome) analysis with the QIIME 2 amplicon distribution (2026.1; renamed to \"qiime2\" in 2026.4) in the correct order: manifest import, cutadapt trim-paired primer removal BEFORE dada2 denoise-paired (trunc-len chosen from the demux quality .qzv), feature-classifier classify-sklearn against a version-matched SILVA 138 or Greengenes2 classifier, and diversity core-metrics-phylogenetic — teaching the .qza\u002F.qzv artifact-and-provenance model and the 2026.1 feature-table summarize change (the former summarize_plus). Use when the request mentions QIIME2, QIIME 2, qiime, 16S, 18S, ITS, amplicon, microbiome, ASV, DADA2 denoising, feature table, taxonomic classification, or core-metrics diversity. For downstream alpha\u002Fbeta diversity, PCoA, and PERMANOVA on the exported feature table prefer alterlab-scikit-bio; this is conda-only (no pip install). Part of the AlterLab Academic Skills suite.'\nlicense: MIT\nallowed-tools: Read Write Edit Bash(python:*) Bash(uv:*) Bash(qiime:*) Bash(conda:*)\ncompatibility: \"Requires the QIIME 2 amplicon conda environment (cannot be pip-installed); commands are run via the `qiime` CLI. Pretrained classifiers and reference data are downloaded from the QIIME 2 Library. The helper scripts in scripts\u002F are stdlib-only and run under `uv run python` without a QIIME 2 env.\"\nmetadata:\n    skill-author: AlterLab\n    version: \"1.0.0\"\n---\n\n# QIIME 2 Amplicon — 16S\u002FITS Microbiome Pipeline (FASTQ → Feature Table → Taxonomy → Diversity)\n\nThe command-line, workflow-runner entry point for marker-gene (amplicon) microbiome\nanalysis. Given raw demultiplexed paired-end reads, it walks the **canonical QIIME 2\norder** — import → primer trim → denoise → classify → diversity — and teaches the two\nthings people get wrong most: **trimming primers BEFORE DADA2**, and the **.qza\u002F.qzv\nprovenance model**. It is the raw-data-to-result pipeline that hands a feature table off\nto in-memory analysis skills (see routing below).\n\nPinned to **QIIME 2 2026.1** (the `amplicon` distribution). Forward-compat note: the\ndistribution is **renamed `qiime2` in 2026.4** — the env name and channel URL change, the\nplugin commands below do not.\n\n## When to Use This Skill\n\nUse this skill when the request involves running an amplicon \u002F microbiome pipeline from\nsequencing reads:\n\n- \"Run a QIIME 2 16S pipeline on my paired-end reads.\"\n- \"I have ITS amplicon FASTQs — denoise with DADA2 and assign taxonomy.\"\n- \"Build a feature table \u002F ASV table and classify against SILVA.\"\n- \"Pick truncation lengths from my quality plot and run core-metrics diversity.\"\n- \"How do I trim primers before DADA2 in QIIME 2?\"\n- \"What's the right order of QIIME 2 commands?\"\n\n### Does NOT Trigger — route these elsewhere\n\n| The request is really about… | Route to |\n|------------------------------|----------|\n| Alpha\u002Fbeta diversity, UniFrac, **PCoA ordination, PERMANOVA** on an already-exported feature\u002Fdistance table (in-memory, Python) | `alterlab-scikit-bio` |\n| Building \u002F manipulating a phylogenetic tree, tree visualization, or comparative phylogenetics outside QIIME 2 | `alterlab-phylogenetics` \u002F `alterlab-etetoolkit` |\n| **Shotgun metagenomics** taxonomic profiling, MAG assembly, functional genes (not marker-gene amplicons) | not in this skill — amplicon only; flag the gap |\n| **RNA-seq** transcript quantification (salmon\u002Fkallisto), differential expression | `alterlab-rnaseq-quant` → `alterlab-pydeseq2` |\n| **Variant calling** FASTQ → VCF (germline\u002Fsomatic) | `alterlab-nf-core-sarek` |\n| Protein\u002Fnucleotide **sequence similarity search** (BLAST+\u002FDIAMOND) | `alterlab-blast` |\n| **Spatial** transcriptomics neighborhood\u002Fenrichment analysis | `alterlab-squidpy-spatial` |\n| Quick one-off gene\u002Fsequence\u002Fdatabase lookups | `alterlab-gget` |\n| Reading\u002Fwriting BAM\u002FSAM\u002FVCF, alignment file surgery | `alterlab-pysam` |\n\nThis skill is **amplicon (marker-gene) only**. If the data is shotgun metagenomic,\nsingle-cell, or anything other than 16S\u002F18S\u002FITS marker-gene sequencing, say so and stop.\n\n## The Artifact Model (.qza \u002F .qzv) — read this first\n\nEverything in QIIME 2 is a **typed, zipped artifact** that records its own provenance:\n\n- **`.qza`** — a QIIME 2 **Artifact**: data (a feature table, sequences, a classifier)\n  plus an embedded **semantic type** (e.g. `SampleData[PairedEndSequencesWithQuality]`,\n  `FeatureTable[Frequency]`) and a full **provenance graph** of every action that\n  produced it.\n- **`.qzv`** — a **Visualization**: a human-viewable report (quality plots, summaries,\n  diversity emperor plots). Drag it into **https:\u002F\u002Fview.qiime2.org** (offline, in-browser)\n  or run `qiime tools view file.qzv`.\n- Provenance is the reproducibility win: any `.qza\u002F.qzv` carries the exact commands,\n  parameters, and plugin versions that made it. Keep artifacts, not just exports.\n\nTreat semantic types as the contract: an action only accepts artifacts of the type it\ndeclares, which is why import (step 1) matters so much.\n\n## The Canonical Order (do not reorder)\n\n```\nmanifest import → cutadapt trim-paired (primers) → dada2 denoise-paired\n   → feature-table summarize → feature-classifier classify-sklearn\n   → phylogeny → diversity core-metrics-phylogenetic\n```\n\n**Primer trimming comes BEFORE DADA2.** DADA2 models per-base error rates; leftover\nprimer\u002Fadapter bases corrupt that error model and inflate spurious ASVs. Trim with\n`cutadapt trim-paired` first, then denoise. (If your reads are already primer-free —\ne.g. EMP-style — you can skip cutadapt, but verify, don't assume.)\n\n### 0. Install \u002F activate the environment (conda only — no pip)\n\nQIIME 2 **cannot be pip-installed**; it ships as a conda environment. For 2026.1\n(verified env files live in `qiime2\u002Fdistributions`):\n\n```bash\n# macOS (Apple Silicon \u002F Intel) — 2026.1 amplicon distribution\nconda env create \\\n  --name qiime2-amplicon-2026.1 \\\n  --file https:\u002F\u002Fraw.githubusercontent.com\u002Fqiime2\u002Fdistributions\u002Fdev\u002F2026.1\u002Famplicon\u002Freleased\u002Fqiime2-amplicon-macos-latest-conda.yml\n# Linux: swap the filename for qiime2-amplicon-ubuntu-latest-conda.yml\nconda activate qiime2-amplicon-2026.1\nqiime info   # confirm version + installed plugins\n```\n\nFor **2026.4**, the official command uses the renamed distribution\n(`--name rachis-qiime2-2026.4`, file `rachis-qiime2-*-conda.yml`); see the QIIME 2\nLibrary quickstart. Full install detail and the env-file matrix:\n[`references\u002Finstallation.md`](references\u002Finstallation.md).\n\n> Bulk DADA2 denoising and classifier training are CPU\u002FRAM heavy. On Cem's M4 Max these\n> run fine locally — keep them off the API and run them in a `conda activate`d shell.\n\n### 1. Import demultiplexed paired-end reads (manifest)\n\nUse a **manifest** (a TSV mapping sample IDs → absolute FASTQ paths) so you control\nexactly which files map to which sample. Format: `PairedEndFastqManifestPhred33V2`\n(verified in `q2-types`).\n\n```bash\nqiime tools import \\\n  --type 'SampleData[PairedEndSequencesWithQuality]' \\\n  --input-format PairedEndFastqManifestPhred33V2 \\\n  --input-path manifest.tsv \\\n  --output-path demux.qza\n\nqiime demux summarize \\\n  --i-data demux.qza \\\n  --o-visualization demux.qzv     # ← READ THIS to choose trunc-len\n```\n\nManifest schema, single-end and EMP variants, and ITS notes:\n[`references\u002Fimport_and_manifest.md`](references\u002Fimport_and_manifest.md).\nGenerate a manifest from a folder of FASTQs with\n[`scripts\u002Fmake_manifest.py`](scripts\u002Fmake_manifest.py).\n\n### 2. Trim primers with cutadapt (BEFORE DADA2)\n\n```bash\nqiime cutadapt trim-paired \\\n  --i-demultiplexed-sequences demux.qza \\\n  --p-front-f GTGYCAGCMGCCGCGGTAA \\   # forward primer (example: 515F)\n  --p-front-r GGACTACNVGGGTWTCTAAT \\  # reverse primer (example: 806R)\n  --p-discard-untrimmed \\\n  --o-trimmed-sequences demux-trimmed.qza\nqiime demux summarize --i-data demux-trimmed.qza --o-visualization demux-trimmed.qzv\n```\n\n`--p-discard-untrimmed` drops reads where the primer was not found (usually what you\nwant for targeted amplicons). Action and flag names verified from the `q2-cutadapt`\nsource. Primer choice by region (515F\u002F806R, ITS1F\u002FITS2, etc.):\n[`references\u002Fpipeline_steps.md`](references\u002Fpipeline_steps.md).\n\n### 3. Denoise with DADA2 → ASVs + feature table\n\nOpen `demux-trimmed.qzv`, read the **interactive quality plot**, and pick truncation\nlengths where median quality drops (forward and reverse independently). Truncated read\nlength must still leave enough overlap to merge pairs.\n\n```bash\nqiime dada2 denoise-paired \\\n  --i-demultiplexed-seqs demux-trimmed.qza \\\n  --p-trunc-len-f 0 --p-trunc-len-r 0 \\   # ← set from the quality .qzv (0 = no truncation)\n  --p-trim-left-f 0 --p-trim-left-r 0 \\\n  --o-representative-sequences rep-seqs.qza \\\n  --o-table table.qza \\\n  --o-denoising-stats denoising-stats.qza\nqiime metadata tabulate \\\n  --m-input-file denoising-stats.qza --o-visualization denoising-stats.qzv\n```\n\nAlways inspect `denoising-stats.qzv`: low merge or chimera-survival rates usually mean\ntrunc-len was too aggressive (no overlap) or primers were not trimmed.\n\n### 4. Summarize the feature table — note the 2026.1 change\n\n```bash\nqiime feature-table summarize \\\n  --i-table table.qza \\\n  --m-sample-metadata-file sample-metadata.tsv \\\n  --o-feature-frequencies feature-frequencies.qza \\\n  --o-sample-frequencies sample-frequencies.qza \\\n  --o-summary table.qzv\n```\n\n**2026.1 breaking change (verified in the release notes):** the old `summarize`\nvisualizer was renamed `_summarize`, and the former **`summarize_plus` pipeline is now\n`summarize`** — so today's `feature-table summarize` *is* the enhanced summary (it also\nemits feature\u002Fsample frequency artifacts). Older tutorials calling `summarize_plus` must\nswitch to `summarize`. Details: [`references\u002Fversion_notes.md`](references\u002Fversion_notes.md).\n\n### 5. Assign taxonomy — VERSION-MATCHED classifier\n\n```bash\nqiime feature-classifier classify-sklearn \\\n  --i-classifier silva-138-99-nb-classifier.qza \\   # MUST match your QIIME 2 version\n  --i-reads rep-seqs.qza \\\n  --o-classification taxonomy.qza\nqiime metadata tabulate --m-input-file taxonomy.qza --o-visualization taxonomy.qzv\n```\n\nA pretrained naive-Bayes classifier is **pickled scikit-learn** — it only loads under the\nQIIME 2 release it was trained on. Download the classifier built for **your** version\nfrom the QIIME 2 Library (SILVA 138 for 16S\u002F18S, Greengenes2 for 16S, UNITE for ITS).\nVersion-match traps and the train-your-own path:\n[`references\u002Fclassifiers.md`](references\u002Fclassifiers.md).\n\n### 6. Phylogeny + core diversity\n\n```bash\nqiime phylogeny align-to-tree-mafft-fasttree \\\n  --i-sequences rep-seqs.qza \\\n  --o-alignment aligned.qza --o-masked-alignment masked.qza \\\n  --o-tree unrooted-tree.qza --o-rooted-tree rooted-tree.qza\n\nqiime diversity core-metrics-phylogenetic \\\n  --i-phylogeny rooted-tree.qza \\\n  --i-table table.qza \\\n  --p-sampling-depth 1103 \\           # ← choose from table.qzv rarefaction; see below\n  --m-metadata-file sample-metadata.tsv \\\n  --output-dir core-metrics\n```\n\n**Sampling depth** is a rarefaction floor: every sample is subsampled to this many reads,\nand samples below it are dropped. Pick it from `table.qzv` to balance depth against sample\nretention — never guess. `core-metrics-phylogenetic` produces Faith's PD, Shannon,\nobserved features, Bray-Curtis \u002F Jaccard \u002F weighted+unweighted UniFrac distance matrices,\nand Emperor PCoA `.qzv`s in one shot.\n\nFor **stats and ordination off the exported table** (PERMANOVA, custom PCoA, alpha\u002Fbeta\nmetrics in Python), export and hand off to **`alterlab-scikit-bio`** — that is the\nin-memory companion to this pipeline.\n\n## Export to hand off downstream\n\n```bash\nqiime tools export --input-path table.qza --output-path exported\u002F   # → feature-table.biom\nqiime tools export --input-path taxonomy.qza --output-path exported\u002F # → taxonomy.tsv\n```\n\n`scripts\u002Fcheck_artifact.py` reads a `.qza\u002F.qzv` (it is just a zip) and prints its semantic\ntype, UUID, and the provenance action list **without a QIIME 2 install** — handy for\nsanity-checking that an artifact is what a downstream step expects.\n\n## Self-Check Before Reporting\n\n- Did primers get trimmed **before** DADA2? If `--p-discard-untrimmed` dropped almost\n  everything, the primer sequences are likely wrong.\n- Were trunc-lens chosen from the **quality `.qzv`**, and does `denoising-stats.qzv` show\n  reasonable merge + non-chimeric retention?\n- Is the classifier **version-matched** to the running QIIME 2 release?\n- Is `--p-sampling-depth` justified from `table.qzv`, not guessed?\n- Did you call `feature-table summarize` (2026.1 = former `summarize_plus`), not a removed\n  action name?\n\n## References\n\n- [`references\u002Finstallation.md`](references\u002Finstallation.md) — conda env files (2026.1 \u002F 2026.4 rename), `qiime info`, why no pip.\n- [`references\u002Fimport_and_manifest.md`](references\u002Fimport_and_manifest.md) — manifest formats, single-end\u002FEMP\u002FITS import.\n- [`references\u002Fpipeline_steps.md`](references\u002Fpipeline_steps.md) — per-step flags, primer sets by region, denoising QC reading.\n- [`references\u002Fclassifiers.md`](references\u002Fclassifiers.md) — SILVA 138 \u002F Greengenes2 \u002F UNITE, version-matching, train-your-own.\n- [`references\u002Fversion_notes.md`](references\u002Fversion_notes.md) — 2026.1 release deltas, 2026.4 `qiime2` rename, `summarize` change.\n\nPart of the AlterLab Academic Skills suite.\n","# QIIME 2 Amplicon — 16S\u002FITS Microbiome Pipeline (FASTQ → Feature Table → Taxonomy → Diversity)\n\nThe command-line, workflow-runner entry point for marker-gene (amplicon) microbiome\nanalysis. Given raw demultiplexed paired-end reads, it walks the **canonical QIIME 2\norder** — import → primer trim → denoise → classify → diversity — and teaches the two\nthings people get wrong most: **trimming primers BEFORE DADA2**, and the **.qza\u002F.qzv\nprovenance model**. It is the raw-data-to-result pipeline that hands a feature table off\nto in-memory analysis skills (see routing below).\n\nPinned to **QIIME 2 2026.1** (the `amplicon` distribution). Forward-compat note: the\ndistribution is **renamed `qiime2` in 2026.4** — the env name and channel URL change, the\nplugin commands below do not.\n\n## When to Use This Skill\n\nUse this skill when the request involves running an amplicon \u002F microbiome pipeline from\nsequencing reads:\n\n- \"Run a QIIME 2 16S pipeline on my paired-end reads.\"\n- \"I have ITS amplicon FASTQs — denoise with DADA2 and assign taxonomy.\"\n- \"Build a feature table \u002F ASV table and classify against SILVA.\"\n- \"Pick truncation lengths from my quality plot and run core-metrics diversity.\"\n- \"How do I trim primers before DADA2 in QIIME 2?\"\n- \"What's the right order of QIIME 2 commands?\"\n\n### Does NOT Trigger — route these elsewhere\n\n| The request is really about… | Route to |\n|------------------------------|----------|\n| Alpha\u002Fbeta diversity, UniFrac, **PCoA ordination, PERMANOVA** on an already-exported feature\u002Fdistance table (in-memory, Python) | `alterlab-scikit-bio` |\n| Building \u002F manipulating a phylogenetic tree, tree visualization, or comparative phylogenetics outside QIIME 2 | `alterlab-phylogenetics` \u002F `alterlab-etetoolkit` |\n| **Shotgun metagenomics** taxonomic profiling, MAG assembly, functional genes (not marker-gene amplicons) | not in this skill — amplicon only; flag the gap |\n| **RNA-seq** transcript quantification (salmon\u002Fkallisto), differential expression | `alterlab-rnaseq-quant` → `alterlab-pydeseq2` |\n| **Variant calling** FASTQ → VCF (germline\u002Fsomatic) | `alterlab-nf-core-sarek` |\n| Protein\u002Fnucleotide **sequence similarity search** (BLAST+\u002FDIAMOND) | `alterlab-blast` |\n| **Spatial** transcriptomics neighborhood\u002Fenrichment analysis | `alterlab-squidpy-spatial` |\n| Quick one-off gene\u002Fsequence\u002Fdatabase lookups | `alterlab-gget` |\n| Reading\u002Fwriting BAM\u002FSAM\u002FVCF, alignment file surgery | `alterlab-pysam` |\n\nThis skill is **amplicon (marker-gene) only**. If the data is shotgun metagenomic,\nsingle-cell, or anything other than 16S\u002F18S\u002FITS marker-gene sequencing, say so and stop.\n\n## The Artifact Model (.qza \u002F .qzv) — read this first\n\nEverything in QIIME 2 is a **typed, zipped artifact** that records its own provenance:\n\n- **`.qza`** — a QIIME 2 **Artifact**: data (a feature table, sequences, a classifier)\n  plus an embedded **semantic type** (e.g. `SampleData[PairedEndSequencesWithQuality]`,\n  `FeatureTable[Frequency]`) and a full **provenance graph** of every action that\n  produced it.\n- **`.qzv`** — a **Visualization**: a human-viewable report (quality plots, summaries,\n  diversity emperor plots). Drag it into **https:\u002F\u002Fview.qiime2.org** (offline, in-browser)\n  or run `qiime tools view file.qzv`.\n- Provenance is the reproducibility win: any `.qza\u002F.qzv` carries the exact commands,\n  parameters, and plugin versions that made it. Keep artifacts, not just exports.\n\nTreat semantic types as the contract: an action only accepts artifacts of the type it\ndeclares, which is why import (step 1) matters so much.\n\n## The Canonical Order (do not reorder)\n\n```\nmanifest import → cutadapt trim-paired (primers) → dada2 denoise-paired\n   → feature-table summarize → feature-classifier classify-sklearn\n   → phylogeny → diversity core-metrics-phylogenetic\n```\n\n**Primer trimming comes BEFORE DADA2.** DADA2 models per-base error rates; leftover\nprimer\u002Fadapter bases corrupt that error model and inflate spurious ASVs. Trim with\n`cutadapt trim-paired` first, then denoise. (If your reads are already primer-free —\ne.g. EMP-style — you can skip cutadapt, but verify, don't assume.)\n\n### 0. Install \u002F activate the environment (conda only — no pip)\n\nQIIME 2 **cannot be pip-installed**; it ships as a conda environment. For 2026.1\n(verified env files live in `qiime2\u002Fdistributions`):\n\n```bash\n# macOS (Apple Silicon \u002F Intel) — 2026.1 amplicon distribution\nconda env create \\\n  --name qiime2-amplicon-2026.1 \\\n  --file https:\u002F\u002Fraw.githubusercontent.com\u002Fqiime2\u002Fdistributions\u002Fdev\u002F2026.1\u002Famplicon\u002Freleased\u002Fqiime2-amplicon-macos-latest-conda.yml\n# Linux: swap the filename for qiime2-amplicon-ubuntu-latest-conda.yml\nconda activate qiime2-amplicon-2026.1\nqiime info   # confirm version + installed plugins\n```\n\nFor **2026.4**, the official command uses the renamed distribution\n(`--name rachis-qiime2-2026.4`, file `rachis-qiime2-*-conda.yml`); see the QIIME 2\nLibrary quickstart. Full install detail and the env-file matrix:\n[`references\u002Finstallation.md`](references\u002Finstallation.md).\n\n> Bulk DADA2 denoising and classifier training are CPU\u002FRAM heavy. On Cem's M4 Max these\n> run fine locally — keep them off the API and run them in a `conda activate`d shell.\n\n### 1. Import demultiplexed paired-end reads (manifest)\n\nUse a **manifest** (a TSV mapping sample IDs → absolute FASTQ paths) so you control\nexactly which files map to which sample. Format: `PairedEndFastqManifestPhred33V2`\n(verified in `q2-types`).\n\n```bash\nqiime tools import \\\n  --type 'SampleData[PairedEndSequencesWithQuality]' \\\n  --input-format PairedEndFastqManifestPhred33V2 \\\n  --input-path manifest.tsv \\\n  --output-path demux.qza\n\nqiime demux summarize \\\n  --i-data demux.qza \\\n  --o-visualization demux.qzv     # ← READ THIS to choose trunc-len\n```\n\nManifest schema, single-end and EMP variants, and ITS notes:\n[`references\u002Fimport_and_manifest.md`](references\u002Fimport_and_manifest.md).\nGenerate a manifest from a folder of FASTQs with\n[`scripts\u002Fmake_manifest.py`](scripts\u002Fmake_manifest.py).\n\n### 2. Trim primers with cutadapt (BEFORE DADA2)\n\n```bash\nqiime cutadapt trim-paired \\\n  --i-demultiplexed-sequences demux.qza \\\n  --p-front-f GTGYCAGCMGCCGCGGTAA \\   # forward primer (example: 515F)\n  --p-front-r GGACTACNVGGGTWTCTAAT \\  # reverse primer (example: 806R)\n  --p-discard-untrimmed \\\n  --o-trimmed-sequences demux-trimmed.qza\nqiime demux summarize --i-data demux-trimmed.qza --o-visualization demux-trimmed.qzv\n```\n\n`--p-discard-untrimmed` drops reads where the primer was not found (usually what you\nwant for targeted amplicons). Action and flag names verified from the `q2-cutadapt`\nsource. Primer choice by region (515F\u002F806R, ITS1F\u002FITS2, etc.):\n[`references\u002Fpipeline_steps.md`](references\u002Fpipeline_steps.md).\n\n### 3. Denoise with DADA2 → ASVs + feature table\n\nOpen `demux-trimmed.qzv`, read the **interactive quality plot**, and pick truncation\nlengths where median quality drops (forward and reverse independently). Truncated read\nlength must still leave enough overlap to merge pairs.\n\n```bash\nqiime dada2 denoise-paired \\\n  --i-demultiplexed-seqs demux-trimmed.qza \\\n  --p-trunc-len-f 0 --p-trunc-len-r 0 \\   # ← set from the quality .qzv (0 = no truncation)\n  --p-trim-left-f 0 --p-trim-left-r 0 \\\n  --o-representative-sequences rep-seqs.qza \\\n  --o-table table.qza \\\n  --o-denoising-stats denoising-stats.qza\nqiime metadata tabulate \\\n  --m-input-file denoising-stats.qza --o-visualization denoising-stats.qzv\n```\n\nAlways inspect `denoising-stats.qzv`: low merge or chimera-survival rates usually mean\ntrunc-len was too aggressive (no overlap) or primers were not trimmed.\n\n### 4. Summarize the feature table — note the 2026.1 change\n\n```bash\nqiime feature-table summarize \\\n  --i-table table.qza \\\n  --m-sample-metadata-file sample-metadata.tsv \\\n  --o-feature-frequencies feature-frequencies.qza \\\n  --o-sample-frequencies sample-frequencies.qza \\\n  --o-summary table.qzv\n```\n\n**2026.1 breaking change (verified in the release notes):** the old `summarize`\nvisualizer was renamed `_summarize`, and the former **`summarize_plus` pipeline is now\n`summarize`** — so today's `feature-table summarize` *is* the enhanced summary (it also\nemits feature\u002Fsample frequency artifacts). Older tutorials calling `summarize_plus` must\nswitch to `summarize`. Details: [`references\u002Fversion_notes.md`](references\u002Fversion_notes.md).\n\n### 5. Assign taxonomy — VERSION-MATCHED classifier\n\n```bash\nqiime feature-classifier classify-sklearn \\\n  --i-classifier silva-138-99-nb-classifier.qza \\   # MUST match your QIIME 2 version\n  --i-reads rep-seqs.qza \\\n  --o-classification taxonomy.qza\nqiime metadata tabulate --m-input-file taxonomy.qza --o-visualization taxonomy.qzv\n```\n\nA pretrained naive-Bayes classifier is **pickled scikit-learn** — it only loads under the\nQIIME 2 release it was trained on. Download the classifier built for **your** version\nfrom the QIIME 2 Library (SILVA 138 for 16S\u002F18S, Greengenes2 for 16S, UNITE for ITS).\nVersion-match traps and the train-your-own path:\n[`references\u002Fclassifiers.md`](references\u002Fclassifiers.md).\n\n### 6. Phylogeny + core diversity\n\n```bash\nqiime phylogeny align-to-tree-mafft-fasttree \\\n  --i-sequences rep-seqs.qza \\\n  --o-alignment aligned.qza --o-masked-alignment masked.qza \\\n  --o-tree unrooted-tree.qza --o-rooted-tree rooted-tree.qza\n\nqiime diversity core-metrics-phylogenetic \\\n  --i-phylogeny rooted-tree.qza \\\n  --i-table table.qza \\\n  --p-sampling-depth 1103 \\           # ← choose from table.qzv rarefaction; see below\n  --m-metadata-file sample-metadata.tsv \\\n  --output-dir core-metrics\n```\n\n**Sampling depth** is a rarefaction floor: every sample is subsampled to this many reads,\nand samples below it are dropped. Pick it from `table.qzv` to balance depth against sample\nretention — never guess. `core-metrics-phylogenetic` produces Faith's PD, Shannon,\nobserved features, Bray-Curtis \u002F Jaccard \u002F weighted+unweighted UniFrac distance matrices,\nand Emperor PCoA `.qzv`s in one shot.\n\nFor **stats and ordination off the exported table** (PERMANOVA, custom PCoA, alpha\u002Fbeta\nmetrics in Python), export and hand off to **`alterlab-scikit-bio`** — that is the\nin-memory companion to this pipeline.\n\n## Export to hand off downstream\n\n```bash\nqiime tools export --input-path table.qza --output-path exported\u002F   # → feature-table.biom\nqiime tools export --input-path taxonomy.qza --output-path exported\u002F # → taxonomy.tsv\n```\n\n`scripts\u002Fcheck_artifact.py` reads a `.qza\u002F.qzv` (it is just a zip) and prints its semantic\ntype, UUID, and the provenance action list **without a QIIME 2 install** — handy for\nsanity-checking that an artifact is what a downstream step expects.\n\n## Self-Check Before Reporting\n\n- Did primers get trimmed **before** DADA2? If `--p-discard-untrimmed` dropped almost\n  everything, the primer sequences are likely wrong.\n- Were trunc-lens chosen from the **quality `.qzv`**, and does `denoising-stats.qzv` show\n  reasonable merge + non-chimeric retention?\n- Is the classifier **version-matched** to the running QIIME 2 release?\n- Is `--p-sampling-depth` justified from `table.qzv`, not guessed?\n- Did you call `feature-table summarize` (2026.1 = former `summarize_plus`), not a removed\n  action name?\n\n## References\n\n- [`references\u002Finstallation.md`](references\u002Finstallation.md) — conda env files (2026.1 \u002F 2026.4 rename), `qiime info`, why no pip.\n- [`references\u002Fimport_and_manifest.md`](references\u002Fimport_and_manifest.md) — manifest formats, single-end\u002FEMP\u002FITS import.\n- [`references\u002Fpipeline_steps.md`](references\u002Fpipeline_steps.md) — per-step flags, primer sets by region, denoising QC reading.\n- [`references\u002Fclassifiers.md`](references\u002Fclassifiers.md) — SILVA 138 \u002F Greengenes2 \u002F UNITE, version-matching, train-your-own.\n- [`references\u002Fversion_notes.md`](references\u002Fversion_notes.md) — 2026.1 release deltas, 2026.4 `qiime2` rename, `summarize` change.\n\nPart of the AlterLab Academic Skills suite.\n",{"name":5,"license":15,"metadata":24,"description":7,"allowed-tools":27,"compatibility":28},{"version":25,"skill-author":26},"1.0.0","AlterLab","Read Write Edit Bash(python:*) Bash(uv:*) Bash(qiime:*) Bash(conda:*)","Requires the QIIME 2 amplicon conda environment (cannot be pip-installed); commands are run via the `qiime` CLI. Pretrained classifiers and reference data are downloaded from the QIIME 2 Library. The helper scripts in scripts\u002F are stdlib-only and run under `uv run python` without a QIIME 2 env.","master","49e49abf2f670d429e0b7e2c0349739c3fdd5cbe","skills\u002Falterlab-qiime2-amplicon",5,"https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium","https:\u002F\u002Fgithub.com\u002Fstanfish06","https:\u002F\u002Fwww.stanfish.dev\u002F","2026-11-14 21:31:37",{"source":8,"skillName":5,"commands":38,"prompt":48,"zipAvailable":49},[39,42,45],{"tool":40,"command":41},"npx","npx skills add stanfish06\u002Fskillquarium --skill alterlab-qiime2-amplicon",{"tool":43,"command":44},"bunx","bunx skills add stanfish06\u002Fskillquarium --skill alterlab-qiime2-amplicon",{"tool":46,"command":47},"pnpm","pnpm dlx skills add stanfish06\u002Fskillquarium --skill alterlab-qiime2-amplicon","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium\u002Fblob\u002Fmaster\u002Fskills\u002Falterlab-qiime2-amplicon\u002FSKILL.md",true,[51,57,61,65,69,73,77,83],{"path":52,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":55,"sourceHtmlUrl":56},"evals\u002Fevals.json","REFERENCE","参考文档",5069,"https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium\u002Fblob\u002Fmaster\u002Fskills\u002Falterlab-qiime2-amplicon\u002Fevals\u002Fevals.json",{"path":58,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":59,"sourceHtmlUrl":60},"references\u002Fclassifiers.md",3343,"https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium\u002Fblob\u002Fmaster\u002Fskills\u002Falterlab-qiime2-amplicon\u002Freferences\u002Fclassifiers.md",{"path":62,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":63,"sourceHtmlUrl":64},"references\u002Fimport_and_manifest.md",2490,"https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium\u002Fblob\u002Fmaster\u002Fskills\u002Falterlab-qiime2-amplicon\u002Freferences\u002Fimport_and_manifest.md",{"path":66,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":67,"sourceHtmlUrl":68},"references\u002Finstallation.md",2923,"https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium\u002Fblob\u002Fmaster\u002Fskills\u002Falterlab-qiime2-amplicon\u002Freferences\u002Finstallation.md",{"path":70,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":71,"sourceHtmlUrl":72},"references\u002Fpipeline_steps.md",4675,"https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium\u002Fblob\u002Fmaster\u002Fskills\u002Falterlab-qiime2-amplicon\u002Freferences\u002Fpipeline_steps.md",{"path":74,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":75,"sourceHtmlUrl":76},"references\u002Fversion_notes.md",2750,"https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium\u002Fblob\u002Fmaster\u002Fskills\u002Falterlab-qiime2-amplicon\u002Freferences\u002Fversion_notes.md",{"path":78,"resourceType":79,"resourceTypeDesc":80,"sizeBytes":81,"sourceHtmlUrl":82},"scripts\u002Fcheck_artifact.py","SCRIPT","脚本",4610,"https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium\u002Fblob\u002Fmaster\u002Fskills\u002Falterlab-qiime2-amplicon\u002Fscripts\u002Fcheck_artifact.py",{"path":84,"resourceType":79,"resourceTypeDesc":80,"sizeBytes":85,"sourceHtmlUrl":86},"scripts\u002Fmake_manifest.py",5204,"https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium\u002Fblob\u002Fmaster\u002Fskills\u002Falterlab-qiime2-amplicon\u002Fscripts\u002Fmake_manifest.py",[88,95,104,111,118,125],{"skillNo":89,"name":90,"nameZh":6,"description":91,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":92,"categories":93,"keywords":94,"publishedAt":19},"01M76ZJTRVA456GD7K8Y3EP82H","zarr-python","Stores and queries chunked N-D scientific arrays with Zarr-Python 3, including codecs, sharding, S3\u002FGCS storage, and NumPy\u002FDask\u002FXarray integration. Use for array layout, bounded I\u002FO, format migration, or scientific metadata preservation.","https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium\u002Fblob\u002Fmaster\u002Fskills\u002Fzarr-python\u002FSKILL.md",[],[],{"skillNo":96,"name":97,"nameZh":6,"description":98,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":99,"licenseFileUrl":100,"sourceHtmlUrl":101,"categories":102,"keywords":103,"publishedAt":19},"01M76ZJQZ15Q1HKC5V85NHE325","yeet","Publish local changes to GitHub by confirming scope, committing intentionally, pushing the branch, and opening a draft PR through the GitHub app from this plugin, with `gh` used only as a fallback where connector coverage is insufficient.","Apache-2.0","https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium\u002Fblob\u002Fmaster\u002Fskills\u002Fyeet\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium\u002Fblob\u002Fmaster\u002Fskills\u002Fyeet\u002FSKILL.md",[],[],{"skillNo":105,"name":106,"nameZh":6,"description":107,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":99,"licenseFileUrl":6,"sourceHtmlUrl":108,"categories":109,"keywords":110,"publishedAt":19},"01M76ZJ8ASZ4QYQ70Y6SBA4Z4S","xarray","N-dimensional labeled arrays and datasets in Python. Built on top of NumPy and Dask. It introduces labels in the form of dimensions, coordinates, and attributes on top of raw NumPy-like arrays, making data analysis in physical sciences more intuitive and less error-prone. Use for working with multi-dimensional scientific data, NetCDF\u002FGRIB\u002FZarr files, climate\u002Fweather\u002Foceanographic datasets, remote sensing, geospatial imaging, large out-of-memory datasets with Dask, and labeled array operations.","https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium\u002Fblob\u002Fmaster\u002Fskills\u002Fxarray\u002FSKILL.md",[],[],{"skillNo":112,"name":113,"nameZh":6,"description":114,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":115,"categories":116,"keywords":117,"publishedAt":19},"01M76ZHFBXD6VXPQ02VCEG202B","writing-mstest-tests","Use when asked to fix, rewrite, update, improve, modernize, show corrected code for, or explain existing MSTest tests or MSTest-specific configuration. Use for \"review\" when corrected code or edits are wanted, even for one pasted assertion or passing tests with bad failure output. Covers expected\u002Factual labels; generic Boolean, collection, string, numeric, null, identity, exception, hard-cast, and object[] checks; TestContext\u002Flifecycle; timeout\u002Fcancellation; OS\u002FCI conditions, retry, cleanup, parallelization, MSTest.Sdk project setup, and MSTESTxxxx. Honor the installed MSTest version. DO NOT USE to design new test cases (code-testing), perform report-only audits, create project files rather than explain MSTest setup, run tests, migrate frameworks, or handle non-MSTest\u002Fnon-.NET code.","https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium\u002Fblob\u002Fmaster\u002Fskills\u002Fwriting-mstest-tests\u002FSKILL.md",[],[],{"skillNo":119,"name":120,"nameZh":6,"description":121,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":122,"categories":123,"keywords":124,"publishedAt":19},"01M76ZDFDF4N88RH6HSYGS7GP2","wgs-prs","End-to-end WGS to polygenic risk score pipeline. Takes paired-end FASTQ files (or a pre-existing VCF) through nf-core\u002Fsarek for variant calling, applies VCF QC (normalisation, hard filtering, Ti\u002FTv and Het\u002FHom checks), then computes polygenic risk scores via the PGS Catalog. Fills the FASTQ to VCF gap upstream of the gwas-prs skill.","https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium\u002Fblob\u002Fmaster\u002Fskills\u002Fwgs-prs\u002FSKILL.md",[],[],{"skillNo":126,"name":127,"nameZh":6,"description":128,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":99,"licenseFileUrl":129,"sourceHtmlUrl":130,"categories":131,"keywords":132,"publishedAt":19},"01M76ZCC1SC68J42QK4MDC8RCG","webapp-testing","Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.","https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium\u002Fblob\u002Fmaster\u002Fskills\u002Fwebapp-testing\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fstanfish06\u002Fskillquarium\u002Fblob\u002Fmaster\u002Fskills\u002Fwebapp-testing\u002FSKILL.md",[],[]]