Taxonomic Coverage Across Databases
This document describes what organisms and clades are covered by different databases in the JGI and KBase lakehouses.
Overview
| Database | Scope | Organisms | Focus |
|---|---|---|---|
| GOLD | Universal | 598K | All sequenced organisms (registry) |
| IMG | Microbial | 287K | Annotated microbial genomes |
| KBase Pangenome | Bacteria/Archaea | 293K | GTDB-classified prokaryotes |
| MycoCosm | Fungi | ~2,500 | Fungal comparative genomics |
| Phytozome | Plants | ~250 | Plant comparative genomics |
| IMG/VR | Viruses | 15.7M | Viral sequences from metagenomes |
| Phage Foundry | Phage-host | ~2,200 | Phage therapy target pathogens |
Data Types
Databases contain a mix of:
| Type | Description | GOLD Count |
|---|---|---|
| Isolate genomes | Cultured organisms | 518K (87%) |
| MAGs | Metagenome-assembled genomes | 79K |
| SAGs | Single-cell amplified genomes | 16K |
| Metagenomes | Community DNA | 223K projects |
| Metatranscriptomes | Community RNA | 40K projects |
GOLD: The Genome Registry
GOLD (Genomes OnLine Database) tracks all sequencing projects at JGI and beyond.
Domain Distribution
| Domain | Organisms | Percentage |
|---|---|---|
| Bacteria | 489,957 | 81.9% |
| Eukarya | 69,683 | 11.6% |
| Viruses | 17,496 | 2.9% |
| Archaea | 6,877 | 1.2% |
| Unknown | 14,048 | 2.4% |
Bacterial Coverage
Top phyla (using modern nomenclature):
| Phylum | Count | % of Bacteria |
|---|---|---|
| Pseudomonadota (Proteobacteria) | 211,920 | 43% |
| Bacillota (Firmicutes) | 139,789 | 29% |
| Actinomycetota (Actinobacteria) | 70,898 | 14% |
| Bacteroidota | 23,577 | 5% |
| Campylobacterota | 11,079 | 2% |
| Cyanobacteriota | 4,828 | 1% |
Top genera:
| Genus | Count | Notes |
|---|---|---|
| Pseudomonas | 24,177 | Environmental/clinical |
| Escherichia | 22,452 | Model organism, pathogens |
| Streptococcus | 22,083 | Human pathogens |
| Staphylococcus | 19,825 | Human pathogens |
| Salmonella | 16,301 | Food safety |
| Bacillus | 12,960 | Industrial/environmental |
| Streptomyces | 12,156 | Natural products |
Bias note: Clinical and industrially-relevant genera are overrepresented.
Archaeal Coverage
| Kingdom | Count | Notes |
|---|---|---|
| Methanobacteriati | 4,358 | Methanogens |
| Thermoproteati | 1,709 | Thermophiles |
| Other | ~800 | Various |
Archaea are underrepresented (1.2%) with roughly equal cultured/uncultured split, unlike bacteria where cultured dominates.
Eukaryotic Coverage
| Kingdom | Count | Notes |
|---|---|---|
| Viridiplantae | 40,978 | Plants (→ Phytozome) |
| Fungi | 21,298 | Fungi (→ MycoCosm) |
| Metazoa | 4,934 | Animals |
| Other protists | ~2,500 | Various |
Viral Coverage
| Realm | Count | Notes |
|---|---|---|
| Orthornavirae | 5,255 | RNA viruses |
| Shotokuvirae | 2,750 | dsDNA viruses (phages) |
| Other | ~9,500 | Various |
Metagenomes and Environmental Samples
Cultured vs Uncultured
| Status | Count | Percentage |
|---|---|---|
| Cultured | 518,241 | 86.7% |
| Uncultured | 79,815 | 13.3% |
Uncultured breakdown:
| Type | Count | Description |
|---|---|---|
| Metagenomic | 63,270 | Binned from metagenomes |
| Single Cell | 15,962 | SAGs |
| Other | ~600 | Enrichments, pooled |
Project Types
| Project Type | Count |
|---|---|
| Genome Analysis (Isolate) | 286,643 |
| Metagenome Analysis | 152,545 |
| Metagenome-Assembled Genome | 78,813 |
| Metatranscriptome Analysis | 29,309 |
| Single Cell Analysis | 14,741 |
Ecosystem Classification
GOLD uses a hierarchical ecosystem classification:
Ecosystem (top level)
└── Ecosystem Category
└── Ecosystem Type
└── Ecosystem Subtype
└── Specific Ecosystem
Top-Level Ecosystems
| Ecosystem | Count | Examples |
|---|---|---|
| Host-associated | 153,878 | Human gut, plant roots |
| Environmental | 80,581 | Soil, ocean, freshwater |
| Engineered | 22,261 | Bioreactors, wastewater |
Host-Associated Breakdown
| Category | Count | Notes |
|---|---|---|
| Human | 85,157 | 55% of host-associated |
| Other Mammals | 21,477 | Livestock, wildlife |
| Plants | 22,816 | Rhizosphere, phyllosphere |
| Birds | 4,804 | Poultry, wildlife |
| Insects | 3,268 | Pollinators, pests |
Human body sites:
| Site | Count |
|---|---|
| Digestive system | 34,767 |
| Respiratory system | 14,679 |
| Circulatory system | 7,645 |
| Skin | ~5,000 |
| Oral cavity | ~4,000 |
Environmental Breakdown
| Type | Count | Examples |
|---|---|---|
| Soil | 31,563 | Forest, agricultural, permafrost |
| Marine | 19,527 | Pelagic, sediment, hydrothermal |
| Freshwater | 7,101 | Lakes, rivers, wetlands |
| Other | ~22,000 | Air, rock, ice |
Special categories: - "Plant-associated metagenome" → Host-associated > Plants - "Soil metagenome" → Environmental > Terrestrial > Soil - "Human gut metagenome" → Host-associated > Human > Digestive system
IMG: Annotated Microbial Genomes
IMG focuses on microbial genomes with comprehensive functional annotation.
Total taxa: 286,758
Domain Distribution
| Domain | Count | Percentage |
|---|---|---|
| Bacteria | 167,336 | 58.5% |
| *Microbiome | 90,371 | 31.6% |
| Viruses | 21,818 | 7.6% |
| Archaea | 4,536 | 1.6% |
| Eukaryota | 1,192 | 0.4% |
| Plasmids | ~1,200 | 0.4% |
Note: "*Microbiome" entries are metagenomic assemblies, not isolate genomes.
Genome Types
IMG distinguishes: - Isolate - Single organism genomes (~173K) - Metagenome - Community assemblies (~90K) - Plasmids - Mobile elements - Genome Fragments - Partial sequences
KBase Pangenome: GTDB-Classified Prokaryotes
The KBase pangenome database contains 293,059 bacterial and archaeal genomes classified using GTDB (Genome Taxonomy Database).
Coverage
| Domain | Genomes | Species Clades |
|---|---|---|
| Bacteria | ~280K | ~80K |
| Archaea | ~13K | ~5K |
GTDB vs NCBI Taxonomy
GTDB provides standardized, rank-normalized taxonomy that often differs from NCBI:
| NCBI Name | GTDB Name | Notes |
|---|---|---|
| Proteobacteria | Pseudomonadota | Renamed |
| Firmicutes | Bacillota | Renamed |
| "Candidatus X" | X | Standardized |
MycoCosm: Fungal Genomes
JGI's fungal genome portal with ~2,500+ genomes across three MySQL servers.
Taxonomic Coverage
| Phylum | Genomes | Examples |
|---|---|---|
| Ascomycota | ~1,500 | Aspergillus, Neurospora, yeasts |
| Basidiomycota | ~800 | Agaricus, Coprinopsis, Laccaria |
| Mucoromycota | ~100 | Mucor, Rhizopus |
| Chytridiomycota | ~50 | Batrachochytrium |
| Other | ~50 | Microsporidia, etc. |
Database Structure
Each fungal genome has its own schema in Dremio:
- myco-db-1 mysql.Aspni7 (Aspergillus niger)
- myco-db-1 mysql.Copci1 (Coprinopsis cinerea)
- myco-db-2 mysql.Lacbi2 (Laccaria bicolor)
Tables per genome: 100-250, including:
| Table | Description |
|---|---|
FilteredModels1 |
Curated gene models (primary gene set) |
GeneCatalog1 |
Gene catalog entries |
InterPro1 |
InterPro domain annotations |
blastp1 |
BLASTP hits |
SM_cluster1 |
Secondary metabolite clusters |
KOG1 |
KOG functional categories |
Naming Convention
Genome schema names follow pattern: {Genus3}{species3}{version}
- Aspni7 = Aspergillus niger v7
- Copci1 = Coprinopsis cinerea v1
- Agabi_varbisH97_2 = Agaricus bisporus var. bisporus H97 v2
Phytozome: Plant Genomes
JGI's plant comparative genomics portal using GMOD Chado schema.
Database: plant-db-7 postgresql.public (306 tables)
Coverage
| Clade | Genomes | Examples |
|---|---|---|
| Angiosperms | ~200 | Arabidopsis, rice, maize, soybean |
| Gymnosperms | ~10 | Pine, spruce |
| Ferns/Lycophytes | ~5 | Selaginella |
| Bryophytes | ~10 | Physcomitrella, Marchantia |
| Algae | ~30 | Chlamydomonas, Volvox, Ostreococcus |
Database Structure
Unlike MycoCosm (schema-per-genome), Phytozome uses a unified Chado schema:
| Table | Description |
|---|---|
feature |
All genomic features (genes, mRNAs, exons) |
featureloc |
Feature coordinates |
organism |
Species metadata |
analysis |
Annotation pipeline metadata |
featureprop |
Feature properties/attributes |
Key Features
- Gene family clustering across all plant genomes
- Expression atlas data
- Synteny and whole-genome duplication analysis
- PFAM/InterPro domain annotations
IMG/VR: Viral Genomes
Database: img-db-1 mysql.imgvr_prod
Total UViGs: 15.7 million viral sequences
Sources
| Source | Count | Description |
|---|---|---|
| Metagenome | 14.2M (90.6%) | Predicted from metagenomes |
| Metatranscriptome | 1.2M (7.3%) | RNA viruses from metatranscriptomes |
| Isolate | 294K (1.9%) | Cultured phages and viruses |
| RefSeq | 14K | Curated reference sequences |
| MAG | 13.5K | From metagenome-assembled genomes |
| SAG | 1.6K | Single amplified genomes |
Viral Realm Distribution
| Realm | Count | Description |
|---|---|---|
| Duplodnaviria | 13.5M (86%) | dsDNA viruses (mostly bacteriophages) |
| Varidnaviria | 1.0M (6.7%) | Large/giant DNA viruses |
| Riboviria | 531K (3.4%) | RNA viruses |
| Unclassified | 521K (3.3%) | No realm assigned |
| Monodnaviria | 72K (0.5%) | ssDNA viruses |
| Adnaviria | 3.7K | Archaeal filamentous viruses |
Key Tables
| Table | Rows | Description |
|---|---|---|
uvig |
15.7M | Main viral sequences |
uvig_lineage |
15.7M | ICTV taxonomy |
uvig_hosts |
1.1M | Predicted hosts |
uvig_domain |
118.7M | Protein domains |
uvig_pfams |
50.4M | PFAM annotations |
Phage Foundry: Phage-Host Interaction Data
The Phage Foundry is a DOE BRaVE (Biopreparedness Research Virtual Environment) initiative led by Vivek Mutalik at Lawrence Berkeley National Laboratory. It focuses on high-throughput characterization of phage-host interactions for phage therapy development against antibiotic-resistant pathogens.
Project Context
| Attribute | Value |
|---|---|
| Funding | DOE BRaVE ($3.6M/year, 2023-2026) |
| Lead Institution | Lawrence Berkeley National Laboratory |
| PI | Vivek Mutalik |
| Focus | Phage therapy for AMR pathogens |
Key Publications: - High-throughput mapping of the phage resistance landscape in E. coli (PLoS Biology, 2020) - BioProject: PRJNA645443 (357 samples, 359 Gbases)
KBase Databases
Strain Modelling Database
Database: phagefoundry_strain_modelling (18 tables)
| Content | Count | Description |
|---|---|---|
| Organisms | 284 | E. coli strains for phage interaction studies |
| Tables | 18 | Experiments, interactions, features, metrics |
Key Tables:
| Table | Description |
|---|---|
strainmodelling_organism |
Bacterial strains (primarily E. coli) |
strainmodelling_genome |
Genome assemblies |
strainmodelling_gene |
Gene annotations with coordinates |
strainmodelling_experiment |
Phage-host interaction experiments |
strainmodelling_interaction |
Phage-host interaction outcomes |
strainmodelling_feature |
Genomic/phenotypic features (receptors, defense systems) |
strainmodelling_protein_family |
Protein family groupings |
E. coli Phage Resistance Landscape Study:
This database contains data from the landmark study mapping host genetic determinants of phage resistance using 14 diverse dsDNA phages:
| Phage | Family | Known Receptor |
|---|---|---|
| T2 | Myoviridae | FadL |
| T4 | Myoviridae | OmpC |
| T5 | Siphoviridae | FhuA |
| T6 | Myoviridae | Tsx |
| T7 | Podoviridae | LPS core |
| λ | Siphoviridae | LamB |
| N4 | Podoviridae | NfrA/NfrB |
| P1vir | Myoviridae | LPS core |
Experimental Approaches: - RB-TnSeq: Random barcode transposon mutagenesis (152K insertions) - CRISPRi: dCas9 transcriptional repression (542 genes screened) - Dub-seq: Dual-barcoded shotgun expression library (GOF screening)
Genome Browser Databases
Schema: phagefoundry_genome_browser (30 tables per organism)
Comparative genome browsers for clinically-relevant bacterial pathogens targeted for phage therapy development:
| Database | Organism | Genomes | Clinical Relevance |
|---|---|---|---|
phagefoundry_acinetobacter_genome_browser |
A. baumannii | 891 | Nosocomial infections, carbapenem-resistant |
phagefoundry_klebsiella_genome_browser_genomedepot |
K. pneumoniae | 220 | Carbapenem-resistant Enterobacteriaceae |
phagefoundry_paeruginosa_genome_browser |
P. aeruginosa | 535 | Cystic fibrosis, burn wounds |
phagefoundry_pviridiflava_genome_browser |
P. viridiflava | 259 | Plant pathogen (bioenergy crops) |
Total: ~1,905 pathogen genomes
Key Tables (per database):
| Table | Description |
|---|---|
browser_genome |
Genome assemblies with NCBI links |
browser_contig |
Contigs with GC content |
browser_gene |
Gene coordinates and types |
browser_protein |
Protein sequences with MD5 hashes |
browser_go_term |
GO functional annotations |
browser_kegg_ortholog |
KEGG pathway mapping |
browser_cog_class |
COG functional categories |
browser_cazy_family |
Carbohydrate-active enzymes |
browser_operon |
Predicted operons |
browser_regulon |
Regulatory networks |
Functional Annotation Coverage: - Gene Ontology (GO) - KEGG Orthologs and Pathways - COG functional classes - CAZy carbohydrate-active enzymes - EC enzyme numbers - eggNOG ortholog descriptions - Transporter Classification (TC)
Cross-References
| External Resource | Link Type |
|---|---|
| NCBI BioProject | PRJNA645443 |
| Figshare | RB-TnSeq, CRISPRi, Dub-seq data |
| NCBI Nucleotide | Genome accessions via external_id |
| IMG/VR | Viral genomes (complementary) |
Example Queries
-- Find E. coli strains with phage interaction data
SELECT o.name, o.full_name, COUNT(i.id) as interactions
FROM phagefoundry_strain_modelling.strainmodelling_organism o
LEFT JOIN phagefoundry_strain_modelling.strainmodelling_interaction i
ON o.id = i.host_id
GROUP BY o.id, o.name, o.full_name
ORDER BY interactions DESC
LIMIT 10;
-- Get genome statistics for Acinetobacter
SELECT name, contigs, size, genes
FROM phagefoundry_acinetobacter_genome_browser.browser_genome
ORDER BY genes DESC
LIMIT 10;
-- Find proteins with specific GO terms (membrane transport)
SELECT p.name, g.go_id, g.name as go_name
FROM phagefoundry_paeruginosa_genome_browser.browser_protein p
JOIN phagefoundry_paeruginosa_genome_browser.browser_protein_go_terms pg
ON p.id = pg.protein_id
JOIN phagefoundry_paeruginosa_genome_browser.browser_go_term g
ON pg.go_term_id = g.id
WHERE g.go_id LIKE 'GO:0055085%' -- transmembrane transport
LIMIT 20;
Cross-Database Organism Lookup
By NCBI Taxonomy ID
Most databases link via NCBI taxonomy:
-- Find organism across databases
SELECT 'GOLD' as db, organism_name
FROM "gold-db-2 postgresql".gold.organism_v2
WHERE ncbi_taxon_id = 562 -- E. coli
UNION ALL
SELECT 'IMG' as db, taxon_display_name
FROM "img-db-2 postgresql".img_core_v400.taxon
WHERE ncbi_taxon_id = 562
By Assembly Accession
For genome-level matching:
-- Match KBase to GOLD via assembly accession
SELECT kb.genome_id, gold.organism_name
FROM kbase_ke_pangenome.genome kb
JOIN "gold-db-2 postgresql".gold.ncbi_assembly gold
ON SUBSTRING(kb.genome_id FROM 4) = gold.assembly_accession
Coverage Gaps and Biases
Overrepresented
| Category | Reason |
|---|---|
| Human pathogens | Clinical relevance, surveillance |
| Model organisms | Research investment |
| Industrial microbes | Biotechnology applications |
| Human gut microbiome | Health research funding |
Underrepresented
| Category | Reason |
|---|---|
| Environmental Archaea | Difficult to culture |
| Deep-sea organisms | Sampling challenges |
| Protists | Genome complexity |
| Non-human host microbiomes | Less funding |
| Rare biosphere | Low abundance |
Data Type Biases
| Bias | Impact |
|---|---|
| Isolate bias (87%) | Missing unculturable diversity |
| Short-read assemblies | Fragmented genomes, missing repeats |
| Reference genome focus | Strain diversity undersampled |
Recommendations
- For prokaryotic diversity: Use KBase pangenome with GTDB taxonomy
- For metagenome context: Use GOLD ecosystem classification
- For fungi: Use MycoCosm (clade-specific annotations)
- For plants: Use Phytozome (clade-specific annotations)
- For viruses/phages: Use IMG/VR for metagenomic viral sequences
- For phage-host interactions: Use Phage Foundry databases (E. coli resistance, AMR pathogens)
- For cross-database queries: Join on NCBI taxonomy ID or assembly accession
Key Tables
GOLD
| Table | Description |
|---|---|
organism_v2 |
Organism metadata with taxonomy |
project |
Sequencing project metadata |
ncbi_taxonomy |
Full NCBI taxonomy |
ncbi_assembly |
Assembly accessions |
IMG
| Table | Description |
|---|---|
taxon |
Genome metadata |
taxon_stats |
Genome statistics |
KBase
| Table | Description |
|---|---|
genome |
GTDB-classified genomes |
gtdb_species_clade |
Species-level groupings |
Phage Foundry
| Table | Database | Description |
|---|---|---|
strainmodelling_organism |
strain_modelling | E. coli strains (284) |
strainmodelling_interaction |
strain_modelling | Phage-host interaction data |
strainmodelling_experiment |
strain_modelling | Experimental conditions/metrics |
browser_genome |
genome_browser | Pathogen genomes (~1,905) |
browser_protein |
genome_browser | Annotated proteins |
browser_go_term |
genome_browser | GO functional annotations |