The Pan-Bacterial Essential Genome

Raw report: essential_genome Feeds into: cofitness-network-architecture, condition-specific-fitness, cross-species-fitness-transferability, evidence-triangulation-for-functional-annotation, experimental-prioritization-of-functional-dark-matter, gene-essentiality, genetic-perturbation-coverage-bias, homology-search-negative-evidence, pangenome-conservation-fitness-decoupling, pangenome-core-boundary-and-clade-size-bias, pangenome-integration, transposon-callability-bias, two-speed-bacterial-genome

Overview

This report analyzes essentiality, orthology, conservation, and function prediction across 221,005 genes from 48 bacteria using Fitness Browser data, bidirectional best-hit (BBH) orthology, pangenome conservation links, and ICA (independent component analysis) fitness modules. It identifies a small set of deeply conserved essential families, quantifies widespread context-dependent essentiality, characterizes poorly annotated orphan essentials, and transfers module-based functional predictions to hypothetical essential genes. ₁

Key Findings

Pan-Bacterial Essential Families

Fifteen gene families were essential in all 48 organisms with no exceptions: the ribosomal proteins rpsC, rplW, rplK, rplB, rplA, rplF, rps11, rpsJ, rpsI, and rpsM; the chaperonin groEL; CTP synthase pyrG; translation elongation factor G fusA; valyl-tRNA synthetase valS; and geranyltranstransferase SelGGPS. ₁

Of 17,222 ortholog families across the 48 organisms, 859 (5.0%) were universally essential, 4,799 (27.9%) were variably essential, and 11,564 (67.1%) were never essential. Among the universally essential families, 839 were strict single-copy families with copy ratio <=1.5 and no non-essential paralogs, while 20 contained paralogs. ₁

Essential Families Overview. From essential_genome.

The analysis identified 41,059 essential genes among 221,005 genes, corresponding to 18.6% of all genes. Essentiality rates ranged from 12.2% in Pedo557 to 29.7% in Magneto. The 2,838,750 BBH pairs yielded 17,222 ortholog groups spanning all 48 organisms. ₁

Essential genes were shorter than non-essential genes, with median lengths of 675 bp and 885 bp, respectively; 17.8% of essential genes were shorter than 300 bp. ₁

Variable Essentiality

Variably essential families had a median essentiality penetrance of 33%; 813 families were more than 50% essential and 704 were less than 10% essential. ₁

The report interprets variable essentiality as evidence that whether a gene is essential depends on genomic context, including paralogs, alternative pathways, and compensatory functions in the accessory genome. The cross-organism result extends the strain-dependent and evolvable essentiality observed within Streptococcus pneumoniae to phylogenetically diverse organisms spanning Proteobacteria, Bacteroidetes, Firmicutes, and Archaea. ₁

The report cites Rosconi et al. (2022), who demonstrated context-dependent essentiality across 36 Streptococcus pneumoniae strains, and says its own work extends this concept from within-species to cross-species variation. In the report’s example, the same gene can be essential in Methanococcus but dispensable in Pseudomonas, depending on metabolic context. ₁

Orphan Essential Genes

The analysis found 7,084 essential genes with no detectable orthologs in any other Fitness Browser organism. Of these orphan essentials, 58.7% were hypothetical, compared with 8.2% of universally essential genes. ₁

There were 8,297 hypothetical essential genes across the 48 organisms, representing 20.2% of all essentials. Of these, 3,912 had orthologs and were predictable by the reported method, while 4,385 were orphans and were not predictable by that method. ₁

Orphan essentials were 49.5% core, whereas universally essential genes were 91.7% core. The report interprets their lower core fraction as evidence that many orphan essentials are strain-specific functions, although it notes that some may represent recently acquired genes, rapidly evolving proteins, or lineage-specific innovations; these explanations are proposed rather than established. The report compares the orphan essentials with Hutchison et al. (2016), who found that 149 of 473 genes (31%) in their minimal Mycoplasma genome had unknown function, and argues that this “dark matter” of essential genes is even larger across diverse bacteria. ₁

Conservation Architecture

Universally essential genes were 91.7% core, variably essential genes were 88.9% core, never-essential genes were 81.7% core, and orphan essential genes were 49.5% core. The corresponding gene counts reported were 6,963 universally essential, 90,844 variably essential, 47,140 never-essential, and 4,683 orphan essential genes. ₁

The report’s key-findings section and its conservation table give figures that do not fully match. The key-findings section states that universally essential genes are 91.7% core versus 80.7% for “non-essential” genes, that 71% of universally essential families are 100% core across all genomes in their species, and that orphan essentials are only 49.5% core. The conservation table instead reports 81.7% core for the “never essential” class of 47,140 genes. The two non-essential figures carry different labels, and the report does not reconcile them. The table also reports 49.5% core for 4,683 orphan essential genes, which differs from the report’s total of 7,084 orphan essentials, and the report does not explain how the table subset relates to the total. ₁

Seventy-one percent of universally essential families were 100% core across all genomes in their species. Essentiality penetrance and core fraction showed a weak positive correlation (rho=0.123, p=1.6e-17): families with more than 80% penetrance were 97.1% core, compared with 92.8% for families with less than 20% penetrance. Clade size did not predict essentiality rate (rho=-0.13, p=0.36). ₁

Conservation Architecture. From essential_genome.

Function Prediction for Hypothetical Essentials

Module transfer from non-essential orthologs in ICA fitness modules generated 1,382 function predictions for hypothetical essential genes, equal to 35.3% of predictable targets. All predictions were family-backed and spanned all 48 organisms. The top predicted functions were TIGR00254 signal transduction, PF00460 flagellar basal body rod, and PF00356 lactoylglutathione lyase. ₁

Prediction counts ranged from 90 predictions for pseudo1_N1B4 to 1 prediction for Caulo. The method uses fitness-module context from non-essential orthologs because essential genes cannot be directly characterized by fitness profiling when disruption is lethal; the transferred context therefore provides indirect functional evidence. ₁

Interpretation and Relevance

The report characterizes the 859 universally essential families as a stringent, experimentally defined core. It says its experimentally defined universally essential set is smaller than computationally conserved sets because the set is restricted to genes whose transposon disruption is lethal across all 48 tested organisms, which it considers more stringent than computational conservation. That wording is in tension with the report’s family-classification table, which defines a universally essential family as essential in every organism where the family has members, so the family need not be present in all 48 organisms. Only 15 families were essential in all 48 organisms. The report does not reconcile the two phrasings. ₁

The report proposes that universally essential families may provide more reliable broad-spectrum antibiotic targets than variably essential families, while emphasizing that essentiality depends on organismal and genomic context. The report states that only the 859 universally essential families are reliable broad-spectrum targets. This is an interpretation from essentiality patterns, not a tested drug-efficacy result. ₁

The report contrasts its approach with Duffield et al. (2010), who used computational prediction to identify 52 conserved essential proteins across 14 bacterial genomes as drug targets. The present work instead uses experimental RB-TnSeq (random barcode transposon sequencing) essentiality data across 48 organisms, and the report presents this as empirical validation of conservation patterns. ₁

The Fitness Browser data analyzed here were generated by Price et al. (2018). The report says it adds a cross-organism essentiality dimension to those data, enabling function prediction for essential genes that fitness profiling cannot reach. ₁

Data Sources and Outputs

Inputs used by the report ₁:

• Bidirectional best BLAST hit pairs: Fitness Browser besthit table.
• Gene-to-cluster conservation mapping: KBase pangenome link table (conservation_vs_fitness/data/fb_pangenome_link.tsv).
• Co-regulated gene modules: ICA fitness modules (fitness_modules/data/modules/).
• Cross-organism module families: fitness_modules/data/module_families/.
• Functional category assignments: SEED annotations (conservation_vs_fitness/data/seed_annotations.tsv).

Generated data files ₁:

• data/all_ortholog_groups.csv contains 179,237 gene-to-ortholog-group assignments across 17,222 ortholog groups.
• data/family_conservation.tsv contains per-family pangenome conservation metrics for 16,758 families with links.

Figures ₁:

• figures/essential_families_overview.png is a 4-panel figure covering family classification counts, size distribution by class, essentiality penetrance and annotation status.
• figures/essential_families_heatmap.png shows essentiality status across 48 organisms for the top 40 universally essential families.
• figures/conservation_architecture.png is a 4-panel figure covering core fraction by class, family-level core distribution, penetrance versus conservation, and clade size versus essentiality.

Caveats and Limitations

The essential-gene definition is an upper bound: genes without transposon insertions may lack insertions because of small size, AT-rich sequence, or scaffold-edge effects rather than true essentiality. ₁

RB-TnSeq, or random barcode transposon sequencing, defines essentiality under specific library-construction conditions, typically rich media. Genes essential only under stress may therefore be missed. ₁

BBH orthology is conservative and can miss paralogs, gene fusions, and distant homologs; consequently, some apparent orphan essentials may have undetected orthologs with diverged sequences. ₁

Connected components can over-merge unrelated genes through transitive connections in the BBH graph, particularly for multi-domain proteins. ₁

The 48-organism set is taxonomically limited and biased toward culturable Proteobacteria, so essentiality patterns in uncultured lineages, Actinobacteria, or Firmicutes are underrepresented. ₁

Module-transfer predictions are indirect because they derive from non-essential orthologs in other organisms, and the function of an essential gene may have diverged from that of its ortholog. ₁

Future Directions

The report proposes characterizing the 7,084 orphan essentials for mobile-element association, recent acquisition, rapid evolution, and functional categories; experimentally testing the 1,382 module-transfer predictions through CRISPRi knockdown under module-informed conditions; linking variable essentiality to metabolic pathway completeness; expanding Fitness Browser taxonomic coverage; and comparing the results with the Database of Essential Genes to identify families that are essential in organisms outside the Fitness Browser. The Database of Essential Genes is a proposed external comparison only; this report did not analyze it. ₁

Slots Into

• gene-essentiality — The 859 universally essential families, 4,799 variably essential families, and 7,084 orphan essentials quantify conserved and context-dependent essentiality across 48 organisms. ₁
• pangenome-integration — Core fractions for universally, variably, never-essential, and orphan essential genes connect essentiality to pangenome conservation architecture. ₁
• cofitness-network-architecture — ICA module transfer produces 1,382 family-backed function predictions for hypothetical essential genes by using fitness context from non-essential orthologs. ₁
• metabolic-model-gapfilling — The proposed analysis of alternative pathways and pathway completeness provides a concrete route for explaining variable essentiality through metabolic context. ₁
• pangenome-conservation-fitness-decoupling — The core fractions by essentiality class, the unreconciled 80.7% versus 81.7% non-essential figures and the weak penetrance–core correlation (rho=0.123) bear on how tightly essentiality tracks conservation. ₁
• pangenome-core-boundary-and-clade-size-bias — Clade size did not predict essentiality rate (rho=-0.13, p=0.36). ₁
• experimental-prioritization-of-functional-dark-matter — The 8,297 hypothetical essentials, including the 4,385 orphans that module transfer cannot predict, define a pool of uncharacterized essential genes. ₁
• cross-species-fitness-transferability — Ortholog-module transfer infers essential-gene functions from fitness data on non-essential orthologs, with the caveat that the functions may have diverged. ₁
• evidence-triangulation-for-functional-annotation — The 1,382 module-transfer predictions are family-backed, indirect functional evidence. ₁
• transposon-callability-bias — Essential-gene calls are an upper bound because short genes, AT-rich sequence and scaffold edges can lack insertions; essential genes are shorter than non-essential genes (median 675 bp vs 885 bp). ₁
• homology-search-negative-evidence — BBH orthology can miss diverged orthologs, which can create apparent orphans, and connected components can over-merge unrelated genes. ₁
• condition-specific-fitness — RB-TnSeq essentiality is defined under library-construction conditions, typically rich media, and misses genes that are essential only under stress. ₁
• genetic-perturbation-coverage-bias — The 48-organism panel is biased toward culturable Proteobacteria. ₁