22,000 Fungal Genomes Still Leave Most Genera Unmapped

More than 22,000 fungal genome assemblies are publicly available, yet the collection represents only about half of fungal orders and roughly one eighth of fungal genera. A new systematic review says the impressive total masks a much narrower biological map—an agricultural problem because crop pathogens, beneficial root partners and possible biocontrol organisms can sit in the missing territory.

The peer-reviewed analysis was published in Nature Genetics on 3 November 2026 and led by Rothamsted Research and the Earlham Institute. The authors assessed public assemblies for taxonomic coverage, quality and usefulness instead of treating the headline count as proof that fungi are well characterized.

One represented genus for about seven missing

Approximately 50% of fungal orders have an assembly, but only about 12.5% of genera do. The approximate one-in-eight share implies that for every represented genus, about seven remain unrepresented. That comparison is an Agrilinkage calculation from the reported fraction, not a direct count in the paper.

Coverage is also concentrated. Ascomycetes account for 79% of available genomes. This group contains yeasts, moulds and important crop pathogens associated with Fusarium head blight, Septoria tritici blotch, rice blast, take-all and powdery mildew. Basidiomycetes—including mushrooms and major rust and smut pathogens—account for about 17%. All remaining fungal phyla together make up only 4%.

Agrilinkage chart from the proportions reported by Kroll and colleagues. Percentages are rounded.

A sequence is not automatically a usable genome map

The review identifies a second bottleneck: annotation. An assembly reconstructs DNA sequence; annotation identifies features such as protein-coding genes. The number of assemblies is rising, but the proportion with good annotation is declining, the authors found.

That limits comparisons. A plant pathologist investigating virulence, resistance-breaking mechanisms or fungicide targets needs equivalent features across isolates and species. Raw sequence with missing or inconsistent annotation makes those comparisons slower and less reliable.

Metadata can be just as important. The review highlights missing information about ploidy and aneuploidy, accessory chromosomes, and whether cells contain one nucleus or multiple genetically distinct nuclei. Fungal genomes can be flexible within a species or life stage; without those details, researchers may confuse species differences with strain structure or chromosome content.

Why farms should care

Disease control increasingly depends on anticipating evolution. Resistant varieties and fungicides impose selection pressure, and pathogen populations respond. Multiple high-quality genomes within a species or genus can reveal which regions are stable, which vary and which accessory elements carry traits that alter virulence or chemical sensitivity.

Sparse sampling weakens that early-warning system. A well-sequenced model species cannot substitute for an under-sampled lineage with different genome organization. It can also make diagnostics riskier: a molecular test built from too narrow a reference set may miss divergent strains or misclassify close relatives.

The blind spot is not only pathogenic. Mycorrhizal fungi support nutrient exchange, soil aggregation and stress responses; other fungi are candidates for biological control, fermentation and new enzymes. Under-representing those branches narrows the pool from which agricultural and food technologies can be developed.

Better metrics, not only more genomes

Standard completeness scores can behave differently across fungal groups. Even highly contiguous telomere-to-telomere assemblies may score lower where a lineage has compact organization or is deeply diverged from reference gene sets. A single threshold can therefore penalize unusual but biologically valid genomes.

A useful global programme must sequence neglected branches, add multiple isolates for agriculturally important species, improve consistent annotation and require richer metadata. Volume alone does not solve those problems.

The reviewed dataset included NCBI and unrestricted Joint Genome Institute assemblies available through 27 October 2025. A supplementary check of later deposits found the sampling biases unchanged, but the 22,000-plus total remains a reviewed snapshot rather than a live census on 4 November 2026.

What the review does—and does not—show

This is a genomic-resource review, not a field trial and not evidence that a particular disease will worsen. It diagnoses research infrastructure: which fungi have been sequenced, how complete and annotated the records are, and whether the collection supports sound comparisons.

Its agricultural value is indirect but substantial. Better references can strengthen diagnostics, resistance monitoring, gene discovery and the study of beneficial fungi. They do not by themselves deliver a resistant crop or new fungicide; those outcomes still require experiments, surveillance and field validation.

Sources and method

Peer-reviewed paper: Kroll et al., “Progress and disparities in fungal genomics in the long-read era,” Nature Genetics (3 November 2026).

Institutional summary: Rothamsted Research, “New global analysis reveals huge gaps in fungal genomics”.

Agrilinkage used the paper metadata and the institutes' quantitative summary. The represented-to-unrepresented genus comparison is derived from approximate one-eighth coverage and is labelled as a calculation.