Compressible but not redundant: knockout recovery and assay-panel architecture of the Neurospora crassa deletion phenome

Classical phenotype assays remain central to functional genetics, but a deletion phenome reflects both assay design and the set of knockouts that can be recovered for phenotyping. We integrated current Neurospora crassa knockout-availability records with a published complete-data phenome of 1,168 deletion strains scored across 10 growth and developmental assays. Among 8,892 genes with interpretable current knockout status, 7,239 had an available homokaryon and 1,653 were represented only as heterokaryons. Within iJDZ836 model genes, Dreyfuss-predicted essential genes were strongly enriched among heterokaryon-only knockouts; iJDZ836 genes as a class were also enriched, showing that the phenotyped matrix lies downstream of a substantial recovery filter. Within the complete phenome, the 10-assay panel resolved 180 discrete profiles. Exact enumeration of all 1,023 non-empty assay subsets showed that 90% of full-panel profile resolution required eight assays and complete recovery required all 10. By contrast, three assays - basal hyphal growth rate, aerial hyphal height, and ascospore number - recovered at least 95% of full-panel normalized mutual information with the published 40-cluster architecture. Serial developmental traits were strongly dependent, but grouping the 10 assays into six biological modules preserved the main result: three modules comprising four assays recovered at least 95% of cluster normalized mutual information, whereas complete profile recovery required all six modules and all 10 assays. Thus, knockout recovery shapes which genes enter the phenome, and within that filtered set broad phenotype architecture is compressible while complete profile resolution is not.