Menu
Skills in this repository
SkillsMP has collected 163 skills from PKU-EMBL/Metagenomics-Skills. Open a skill to review its source and details.
PKU-EMBL/Metagenomics-SkillsShowing 40 of 163 collected skills.
Detect anti-phage defense systems in prokaryotic genomes. Use this skill when working with defensefinder: installing, choosing parameters, running commands, interpreting outputs, or troubleshooting. Upstream: https://github.com/mdmparis/defense-finder. For…
Route biogeochemical / elemental cycling annotation for shotgun metagenomics and MAGs: carbon (C), nitrogen (N), sulfur (S), iron (Fe), phosphorus (P), hydrogen (H2/H), oxygen (O), methane (CH4), and other metals/metalloids (Mn, As, Se, …) via METABOLIC,…
FeGenie — HMM-based identification of iron-related genes and neighborhoods in genomes/metagenome assemblies (acquisition, siderophores, Fe redox, storage, magnetosome). Use for Fe cycling potential on contigs or MAGs. Upstream:…
IntegronFinder — detect integrons (intI, attC, cassettes) in bacterial genomes/assemblies, including AMR cassette context. Use for integron-focused MGE mining. Upstream: https://github.com/gem-pasteur/Integron_Finder. Route via microbial-mining; pair with rgi…
ISEScan — automated annotation of insertion sequence (IS) elements in prokaryotic genomes via profile HMMs. Use for IS family catalogues on MAGs or contigs as part of MGE refinement beyond geNomad. Upstream: https://github.com/xiezhq/ISEScan. Route via…
MEBS (Multigenomic Entropy-Based Score) — score genomes/metagenomes for biogeochemical cycle machinery via informative Pfam domains; strongest for sulfur (S), also supports N, O, CH4, and Fe cycle scores. Use for comparative cycle enrichment across large…
METABOLIC (AnantharamanLab) — high-throughput metabolic and biogeochemical trait profiling of MAGs/SAGs/isolates with C, N, S, and other elemental cycle diagrams (METABOLIC-G genomes-only; METABOLIC-C with reads / community networks). Use for multi-element…
mobileOG-db — manually curated protein families for bacterial MGE life-cycle functions (integration/excision, replication/recombination/repair, stability/defense, transfer, phage-related) with diamond + mobileOG-pl parsing. Use to annotate MGE machinery on…
PADLOC (Prokaryotic Antiviral Defence LOCator) — HMM + synteny detection of antiviral defence systems (CRISPR-Cas and many other systems) in prokaryotic genomes. Use as a complement or alternative to DefenseFinder for defenseome surveys. Upstream:…
QSAP — quorum-sensing (QS) related gene annotation pipeline against the QSP database (DIAMOND and/or HMMER). Use when mining autoinducer / QS circuitry (AHL, AIP, AI-2-related families as covered by QSP) from proteins or assemblies. Upstream:…
ESM-2 family protein language models (ESMFold / ESM Atlas). Use when: (1) Amino-acid embeddings or MLM scores pinned to ESM-2, (2) Single-sequence structure context (ESMFold) or ESM Metagenomic Atlas via Foldseek, (3) Backbone embeddings for PLMSearch /…
ESM C / ESM Cambrian (ESMC) — next-generation protein representation language models scaling beyond ESM-2 (300M / 600M open weights; larger tiers may need Forge/API). Use when: (1) Stronger AA embeddings than ESM-2 for metagenomic proteins, (2) Representation…
Route genomic/protein language models and Steinegger-lab structure/sequence search for shotgun metagenomics. Use when: (1) Choosing gLMs vs PLMs vs classical homology, (2) MMseqs2 vs Foldseek vs PLM search, (3) Spacedust / AlphaGEM, (4) Distinguishing…
ProstT5 — bilingual protein language model translating amino-acid sequences and Foldseek 3Di structure tokens (fine-tuned from ProtT5). Use when: (1) Building Foldseek DBs from FASTA via --prostt5-model, (2) Sequence-only structure-aware search without…
Route protein language models (PLMs / pLMs) for metagenomics: representation models (ESM-2, ESM C / ESMC), structure-aware SaProt (Westlake), bilingual AA↔3Di (ProstT5 / ProtT5), remote-homology search (PLMSearch, TM-Vec/DeepBLAST), and PLM phylogeny…
Route shotgun metagenomics work to the right public databases, catalogues, and benchmark datasets. Use when: (1) Choosing SRA/ENA vs MGnify vs GTDB vs proGenomes vs CAMI, (2) Distinguishing raw archives, genome warehouses, biome catalogues, tool indexes, and…
SaProt — Westlake University structure-aware protein language model using an AA+3Di (Foldseek) structure-aware vocabulary (ICLR 2024). Use when: (1) Structure-informed PLM embeddings for function / fitness / variant tasks, (2) Comparing against ESM-2 with…
TM-Vec and DeepBLAST — deep learning for protein remote homology and structural alignment from sequences (predict TM-scores; align homologous regions without solved structures). Use when: (1) Searching structure-like similarity in large sequence DBs, (2)…
Route genome language models (gLMs) for metagenomics and microbial genomics: representation models (Nucleotide Transformer / NTv3, DNABERT-2, Caduceus), generative models (Evo / Evo 2), and mixed-modality metagenomic LMs (OMG + gLM2). Use when choosing DNA…
Route all homology / similarity search paradigms for shotgun metagenomics, including the Steinegger/Söding stack (MMseqs2, HH-suite, Foldseek family, PLASS, MetaEuk, Petasearch), vector DB (ERAST), and genome warehouse search (Phylign, LexicMap). Use when…
LucaPhylo — alignment-free phylogenetic inference / placement via hyperbolic protein language models (ESM2 + hyperbolic geometry). Use when: (1) MSA-based trees fail on highly divergent proteins, (2) Exploring PLM phylogeny as a complement to…
PLMSearch — protein language model embeddings for fast, sensitive remote homology search from sequences alone. Use when: (1) Homology is too remote for BLAST/MMseqs, (2) Structure search is unavailable but structural similarity is expected, (3) Screening…
Caduceus — bi-directional equivariant long-range DNA language model (Mamba / RC-aware) for long genomic contexts. Use when: (1) Long DNA sequences where transformer gLMs are costly, (2) Reverse-complement equivariant embeddings (Caduceus-PS) or RC-augmented…
DNABERT-2 — efficient multi-species genome language model (BPE + ALiBi) with the Genome Understanding Evaluation (GUE) benchmark. Use when: (1) DNA embeddings or fine-tunes on multi-species tasks, (2) Comparing representation gLMs without NT/Evo-scale…
Evo 2 — long-context DNA language model for genome modeling and design across all domains of life (Arc Institute; StripedHyena 2 / OpenGenome2). Use when: (1) Autoregressive DNA likelihoods or embeddings at up to ~1 Mb context, (2) Sequence generation /…
InstaDeep Nucleotide Transformer family — DNA foundation models from NT/NT-v2 through NTv3 (multi-species, long-context seq–function and generation). Use when: (1) DNA embeddings or MLM scores on contigs/MAGs, (2) Functional-track / genome-annotation style…
Foldcomp — compress and index large protein structure sets (Steinegger lab) for efficient storage and random access before Foldseek-scale searches. Use when: (1) Archiving AFDB-scale PDB collections, (2) Reducing disk for structure search pipelines, (3)…
FoldMason — multiple protein structure alignment at scale (Steinegger lab). Use when: (1) Building structural MSAs from AFDB / predicted models, (2) Comparing folds across MAG protein families after Foldseek clustering, (3) Preparing structure-aware…
HH-suite3 (HHblits / HHsearch) — remote protein homology via HMM–HMM alignment (Söding lab). Use when: (1) Sequence identity is too low for BLAST/MMseqs easy hits, (2) Building/searching profile HMMs (BFD, UniClust, Pfam-style DBs), (3) Feeding sensitive MSAs…
Unicore — universal structure-based core-gene phylogeny using Foldseek and ProstT5 (Steinegger lab). Use when: (1) Building phylogenies from proteomes via structural core genes, (2) 3Di MSAs for tree inference, (3) Scaling beyond sequence-only marker sets.…
MetaEuk — sensitive gene discovery and annotation on eukaryotic metagenomic contigs using MMseqs2 homology + exon dynamic programming (Söding lab). Use when: (1) Calling genes on eukaryotic contigs/MAGs, (2) Annotating via protein or profile targets, (3)…
Petasearch — efficient parallelized peta-scale protein database search (Steinegger lab). Use when: (1) Searching enormous protein collections beyond routine MMseqs2 setups, (2) Preindexing large target lists for repeated queries, (3) Selective Smith–Waterman…
Route phylogenetic and evolutionary analyses for shotgun metagenomics and MAGs: GTDB-Tk placement, genome/marker trees (PhyloPhlAn, IQ-TREE, FastTree), alignment-free PLM phylogeny (LucaPhylo), and ANI species boundaries (skani/FastANI). Use when building or…
PLASS — protein-level assembler that recovers many more protein sequences from metagenomic reads than nucleotide assembly alone (Söding / Steinegger). Same package ships PenguiN (protein-guided nucleotide assembly). Use when: (1) Maximizing protein catalogue…
ERAST — efficient retrieval-augmented homology search combining sequence language-model embeddings with a billion-scale vector database (protein and nucleotide). Use when: (1) Classical BLAST/MMseqs/Foldseek scale poorly, (2) You need vector-DB…
FastTree — approximately-maximum-likelihood trees for large alignments. Use when: (1) Draft phylogenies from big MSAs, (2) Quick topology screens before IQ-TREE, (3) Resource-limited exploratory MAG trees. Upstream: http://www.microbesonline.org/fasttree/.…
Assign bacterial/archaeal taxonomy to MAGs with GTDB. Use this skill when working with gtdbtk: installing, choosing parameters, running commands, interpreting outputs, or troubleshooting. Upstream: https://github.com/Ecogenomics/GTDBTk. For routing across…
IQ-TREE 2 — fast maximum-likelihood phylogenetic inference with model selection and ultrafast bootstrap. Use when: (1) Building publication-grade trees from marker or concatenated MSAs, (2) Following PhyloPhlAn / custom alignments for MAG phylogenomics, (3)…
LexicMap — efficient nucleotide alignment of moderate-length queries (genes, plasmids, long reads; typically >250 bp) against up to millions of prokaryotic genomes. Use when: (1) Searching genes/plasmids vs huge genome databases, (2) Scaling beyond…
Phylign — BLAST-like alignment of genes, plasmids, or reads against phylogeny-compressed microbial genome collections (e.g. pre-2019 661k bacteria) on ordinary desktops. Uses phylogenetic compression (MiniPhy for building compressed assemblies/COBS indexes;…