| name | tool-selection |
| description | Choose shotgun tools by stage: QC, assembly, mapping, binning, taxonomic annotation, functional annotation, microbial mining, and research analysis (stats/figures). Use when deciding which skill to load — mappers, assemblers, ensemble binning (BASALT recommended), profilers, miners, or DA/plotting tools. For stage order use metagenomics-workflow. For mining use microbial-mining; for stats/figures use research-analysis.
|
| license | MIT |
| category | orchestration |
| tags | ["guidance","tool-selection","mapping","binning","mining","research-analysis","workflow"] |
Tool Selection
Start from the stage, then the deliverable.
By stage
1 QC fastp · kneaddata · fastqc · cutadapt · trimmomatic
2 Assembly megahit · metaspades · metaflye · metamdbg (+ quast)
polish: racon (long) · pilon (short)
coassembly planning: binchicken (low-abundance / novelty)
3 Mapping strobealign | bowtie2 | bwa-mem2 | bwa | minimap2 · lexicmap · phylign · samtools · diamond
4 Binning metabat2 · maxbin2 · concoct · semibin2 · vamb · aamb · taxvamb · comebin · metabinner · lorbin
→ basalt ★ · metawrap · dastool → anvio / deepurify → drep · skani · fastani · coverm · checkm2 · gunc
5 Taxonomy metaphlan · singlem · kraken2+bracken · motus · sylph · metabuli · kaiju · gtdbtk
phylogenomics → phylophlan · iqtree · fasttree · lucaphylo
6 Function humann · prodigal · bakta · prokka · cd-hit · mmseqs2 · salmon
· eggnog-mapper · rgi · homology-search
7 Mining microbial-mining → antismash · bigscape · genomad · virsorter2 · checkv · vcontact
· mobileog · isescan · integronfinder · mob-suite · plasmaag · macrel
· cge-finders · plasmidfinder · resfinder · pointfinder
· disinfinder · virulencefinder · mobileelementfinder
· pmlst · mlst · staramr · abricate
· element-cycling · metabolic · mebs · fegenie · dom-biosurfactant
· hadeg · biosurfdb · dram · dbcan · gapseq · vmh · deeparg
· defensefinder · padloc · qsap
8 Report research-analysis → multiqc · microeco · phyloseq · visualization
· maaslin2 · ancombc · lefse · stamp · sparcc · spieceasi
· nextflow · snakemake
Mapping head-to-head
| Need | Prefer |
|---|
| Fast Illumina → large contigs | strobealign |
| Lab Bowtie2 / KneadData ecosystem | bowtie2 |
| BWA-MEM compatible + speed | bwa-mem2 |
| Long reads | minimap2 |
| Gene/plasmid vs millions of genomes | lexicmap |
| Phylogeny-compressed 661k-style search | phylign |
Taxonomy head-to-head
| Need | Prefer |
|---|
| Marker SGBs / pathways with HUMAnN | metaphlan |
| Unknowns / environments lacking genomes | singlem |
| Fast DNA k-mer + abundance | kraken2 + bracken |
| Joint DNA+AA classification | metabuli |
| Protein-level sensitivity | kaiju |
| Sketch containment / coverage-adjusted ANI | sylph |
| MAG taxonomy | gtdbtk |
ANI head-to-head
| Need | Prefer |
|---|
| Large MAG catalogues / incomplete genomes | skani |
| Literature-compatible FastANI Methods | fastani |
| Dereplication workflow | drep (± ANI backend) |
Mining head-to-head
| Goal | Prefer |
|---|
| BGCs / natural products | antismash → bigscape |
| Resistome | rgi (± deeparg); isolates resfinder / staramr |
| Plasmids / viruses / MGEs | genomad (± virsorter2 → checkv → vcontact) |
| MGE proteins / IS / integrons | mobileog / isescan / integronfinder |
| CGE Finders (+ pMLST / mlst / ABRicate) | cge-finders |
| Plasmid MOB / replicon typing | mob-suite / plasmidfinder (± pmlst) |
| Plasmid genome recovery | plasmaag |
| AMPs | macrel |
| Element cycling (C/N/S/Fe/P/…) | element-cycling |
| DOM / biosurfactant / HC biodegradation | dom-biosurfactant |
| Distilled metabolism | dram |
| CAZymes / glycan DOM substrates | dbcan |
| Metabolic models | gapseq |
| Defense systems | defensefinder / padloc |
| Quorum sensing | qsap |
Ensemble binning
| Skill | When |
|---|
basalt | Recommended multi-assembly refinement |
metawrap | Packaged multi-binner suite |
dastool | Aggregate contig2bin tables |
Single-binner diversity for ensembles: metabat2 · semibin2 · vamb · aamb ·
taxvamb · comebin · metabinner · lorbin (long-read).
Post-bin polish: anvio (interactive) · deepurify (decontaminate) → re-QC.
Coverage for binning
| Practice | Prefer |
|---|
| Related samples available | Multi-coverage BAMs into binners |
| Only one sample | Single-coverage (acknowledge hidden contamination risk) |
| Huge public cohorts / rare taxa | binchicken targeted coassembly → then bin |
Homology search (all paradigms)
| Need | Prefer |
|---|
| Route any homology search | homology-search |
| Sequence / NR / contig tax | mmseqs2 · diamond |
| Profile HMM remote homology | hh-suite |
| Protein-level assembly from reads | plass |
| Eukaryotic gene call | metaeuk |
| Structure / motif / MSA | foldseek · folddisco · foldmason |
| Structure archive | foldcomp |
| Peta-scale protein search | petasearch |
| Vector DB + embeddings | erast |
| Phylogeny-compressed genomes | phylign |
| Gene vs million genomes | lexicmap |
Phylogenomics / trees
| Need | Prefer |
|---|
| Route trees / evolution | phylogenomics |
| MAG taxonomy placement | gtdbtk |
| MAG/genome phylogeny workflow | phylophlan |
| Structural core-gene phylogeny | unicore |
| Publication ML tree | iqtree |
| Fast draft tree | fasttree |
| Alignment-free PLM phylogeny | lucaphylo |
Genome language models
| Need | Prefer |
|---|
| Route genome language models | genome-language-model |
| Metagenomic corpus + gLM2 | omg |
| Nucleotide Transformer / NTv3 | nucleotide-transformer |
| Multi-species efficient gLM (GUE) | dnabert2 |
| Long-range RC-aware DNA LM | caduceus |
| Generative long-context DNA (Evo 2) | evo2 |
Protein language models
| Need | Prefer |
|---|
| Route protein language models | protein-language-model |
| ESM-2 embeddings (pinned / Atlas-era) | esm |
| ESM C / Cambrian embeddings | esmc |
| Structure-aware AA+3Di PLM (Westlake) | saprot |
| AA ↔ 3Di / Foldseek from FASTA | prostt5 (± foldseek) |
| Remote protein homology (PLM search) | plmsearch / deepblast |
| Alignment-free PLM phylogeny | lucaphylo |
AI / LM & structure search (Steinegger + PLMs)
| Need | Prefer |
|---|
| Route LM / Foldseek stack | metagenomics-llm |
| Protein LM detail menu | protein-language-model |
| Conserved gene clusters (structure-aware) | spacedust |
| GEM + structure/dark metabolism | alphagem |
Research analysis (stats · figures · reports · workflows)
| Need | Prefer |
|---|
| Route stats / figures / QC / workflows | research-analysis |
| Diversity + composition R workflow | microeco |
| phyloseq data class / classic plots | phyloseq |
| Multivariable differential abundance | maaslin2 |
| Compositional bias-corrected DA | ancombc |
| Classic LDA biomarkers | lefse |
| GUI exploratory profile stats | stamp |
| SparCC correlations | sparcc |
| SPIEC-EASI networks | spieceasi |
| Aggregated QC HTML | multiqc |
| Figure checklist by stage | visualization |
| Nextflow / nf-core pipelines | nextflow |
| Snakemake rule DAGs | snakemake |
Never
- Skip mapping when binning needs coverage
- Merge taxonomy methods unlabeled
- Treat mining hits as validated products/phenotypes
- Call bins MAGs before QC filters
- Run group statistics before inspecting batch / library-size structure
Public data (short)
| Need | Skill |
|---|
| Raw reads | sra-ena |
| Biome MAG catalogues | mgnify |
| GTDB taxonomy | gtdb / gtdbtk |
| Annotated genome warehouse | progenomes |
| Gold-standard benchmarks | cami (± insilicoseq) |
| Marker profiling (unknowns) | singlem |
| Gene vs million genomes | lexicmap |
| KEGG KO (licensed) | kegg |
| Full map | public-databases |
Related skills
metagenomics-workflow · microbial-mining · homology-search · phylogenomics ·
genome-language-model · protein-language-model · metagenomics-llm ·
element-cycling · dom-biosurfactant · cge-finders · research-analysis ·
study-design ·
public-databases
