Two-Speed Bacterial Genomes Separate Conserved Metabolism from Accessory Innovation

The two-speed bacterial genome model proposes that bacterial genomes contain a conserved core devoted largely to metabolism and housekeeping, alongside a more variable accessory component enriched for mobile, defense, and unknown functions. This page synthesizes the cross-species COG analysis reported in COG Functional Category Analysis. ₁

Evidence from Cross-Species COG Profiles

Clusters of Orthologous Groups (COG) categories were compared across 32 species spanning 9 phyla and 357,623 genes. ₁ The analysis found a consistent functional separation between core genes and novel or singleton genes, supporting the broader interpretation of pangenome-integration that pangenome variation has functional structure rather than being uniformly distributed across genes. ₁

Novel or singleton genes were enriched in COG L, the category for replication, recombination, and mobile elements, by +10.88%, with 100% consistency across species. ₁ This was the strongest reported signal and supports the hypothesis that mobile genetic processes are a major source of accessory genomic novelty. ₁

Novel or singleton genes were also enriched in COG V, defense mechanisms, by +2.83%, with 100% consistency, and in COG S, unknown function, by +1.64%, with 69% consistency. ₁ These enrichments associate accessory novelty with ecological interaction, defense, and incompletely characterized functions, linking this model to genomic-under-representation and phage-defense-syndromes-and-arms-race. ₁

Core genes were consistently enriched in COG J, translation. The report gives this as -4.65% enrichment in the novel-versus-core contrast, with 97% consistency, and calls it the strongest depletion from the novel or singleton class. ₁ Core genes were also enriched in COG F, nucleotide metabolism (reported as -2.09% enrichment, 100% consistency); COG H, coenzyme metabolism (-2.06%, 97% consistency); COG E, amino acid metabolism (-1.81%, 81% consistency); and COG C, energy production (-1.75%, 88% consistency). In each case the negative sign marks a category that is under-represented among novel genes and so enriched in the core. ₁

The central discoveries digest records the same comparison of 32 species across 9 phyla (357,623 genes). It lists singleton genes as enriched for L (+10.88%, 100% consistency), V (+2.83%, 100% consistency), and S (+1.64%, 69% consistency). It lists core genes as enriched for J (-4.65%, 97% consistency, labeled the strongest depletion), F (-2.09%, 100% consistency), H (-2.06%, 97% consistency), E (-1.81%, 81% consistency), and C (-1.75%, 88% consistency). The negative values are the digest’s reported novel-versus-core enrichment values. ₂ The digest restates the cog_analysis result rather than replicating it independently, so it adds no statistical weight to the pattern. _2,1

Taken together, these distributions support an interpretation in which core genes form an ancient, conserved metabolic engine centered on translation, energy production, and biosynthesis, whereas novel genes represent more recent acquisitions associated with ecological adaptation, mobile elements, defense, and niche-specific functions. ₁ The essential-genome analysis supports this conserved-core interpretation: universally essential genes were 91.7% core, compared with 49.5% for orphan essential genes, indicating that experimentally indispensable functions can be either deeply conserved or lineage-specific. ₃ It also refines the model by showing that essentiality is not equivalent to core status: 4,799 ortholog families were variably essential, and 7,084 essential genes had no detectable orthologs in the tested organisms. ₃

The conservation-fitness synthesis supports the coordinated-core part of this interpretation: independent component analysis (ICA), a decomposition method for identifying coordinated fitness modules, identified 1,116 modules across 32 organisms, with 86% core genes versus an 81.5% baseline (odds ratio 1.46, p=1.6e-87), and 59% of modules were more than 90% core genes. ₄ This refines the model by showing that conserved architecture is organized into coordinated functional units, not only isolated housekeeping genes. ₄ The new pangenome linkage analysis supports this association between indispensability and conservation: among 33 organisms, essential genes were 86.1% core versus 81.2% for non-essential genes, with a median odds ratio of 1.56; 18 of 33 organisms showed significant enrichment after Benjamini-Hochberg false discovery rate correction. ₅ The result is modest rather than absolute, consistent with the existing evidence that essentiality and core status are related but not equivalent. ₅

This interpretation complements subsurface-bacillota-specialization by distinguishing the functional character of conserved and variable genome fractions rather than treating genome size alone as the relevant axis. ₁ The conservation-fitness analysis further refines the accessory side of the model: essential-auxiliary genes were 38.2% hypothetical and essential-unmapped genes were 44.7% hypothetical, indicating that some indispensable, lineage-specific functions remain poorly characterized rather than belonging only to a well-annotated conserved core. ₅ The same analysis counted 3,683 essential-auxiliary genes, which are essential for viability but not present in all strains. Of these, 13.4% were enzymes, and their top subsystems were ribosomes, DNA replication, type 4 secretion, and plasmid replication. The report tentatively reads this as strain-specific variants of core machinery plus mobile genetic elements. ₅ This refines the two-speed picture: some indispensable genes sit outside the strict core, and they include both core-like machinery and mobile-element functions. ₅

The same synthesis refines the contrast between conserved core and accessory innovation: core genes were more often burdensome in laboratory conditions than accessory genes, with 24.4% showing positive fitness when deleted versus 19.9% of accessory genes, and core genes were 1.78x more likely to have strong condition-specific phenotypes. ₄ The new Core Gene Paradox — Why Are Core Genes More Burdensome? supports and sharpens this qualification: core genes were more burdensome in Protein Metabolism, Motility, and RNA Metabolism, but the direction reversed for Cell Wall genes, where non-core genes were more burdensome; the reported differences were +6.2, +7.8, +12.9, and -14.1 percentage points, respectively. ₆ Thus, conservation does not imply that the core is uniformly cheap or phenotypically inert; laboratory cost may reflect functions retained because they are advantageous in environments not represented by the assay. ₄

Conservation Gradients Across Functional Classes

Protein-structure annotation supports the enrichment of unknown function in the accessory fraction. Hypothetical rates were 13.8% among 7,095,643 auxiliary-plus-singleton clusters, 11.6% among 5,384,900 auxiliary non-singleton clusters, and 3.8% among 25,571,299 core clusters. ₇ This extends the COG S signal to a much larger cluster set and connects to core-gene-annotation-paradox. _7,1

Genome-scale metabolic reconstruction within a single genus supports the conserved-metabolism side of the model. Of 1,330 unique metabolic reactions, 1,248 (94%) were shared across all 14 Acinetobacter genomes, 62 were variable (present in 2-13 genomes), and only 20 were genome-unique. ₈ Gapfilling accounted for 7.7% of reactions on average, and 243 missing functions were cataloged. Part of this reaction-level core may therefore reflect model completion rather than annotated genes, and the result covers only one genus (see metabolic-model-gapfilling). ₈

Antimicrobial resistance (AMR) genes support the accessory-defense side of the model at pangenome scale. Only 30.3% of AMR clusters were core, against a 46.8% baseline (OR=0.49, chi-squared=23,117, p≈0). The auxiliary genome was 2.2x enriched for AMR (33.6% vs 15.3%). ₉ The depletion held in a paired test of 4,252 species, each with ≥5 AMR clusters: 63.7% showed AMR as less core than their species baseline (Wilcoxon p=1.1e-130, mean difference -0.102). ₉

The same atlas refines this result by separating out intrinsic resistance. It describes beta-lactamases such as ampC, efflux systems such as emhABC, and rifampin monooxygenases such as rox as core-genome residents present in >95% of genomes within a species. ₉ The report also states that these genes are vertically inherited and impose negligible fitness costs. Those statements are interpretations rather than measurements. Defense-related function is therefore not uniformly accessory; see antimicrobial-resistance-fitness-cost. ₉

Metal fitness genes show a graded conservation pattern that supports an axis from pleiotropy to specialization within the model. Fully-core proportions declined across three groups at 57.2%, 50.4%, and 45.7%, an 11.5 percentage-point spread. The groups were general-stress genes, metal-shared genes (important for ≥2 metals), and metal-specific genes (important for exactly 1 metal). ₁₀

A pooled comparison qualifies that gradient (see shared-stress-versus-stressor-specific-fitness). Metal-specific genes were 84.8% core versus 90.2% for general sick genes (Cochran-Mantel-Haenszel p=0.011), but both exceeded the 79.8% baseline. The result therefore only partially supports a split between a core general stress response and accessory specialized resistance, and the accessory genome contributed only modestly to metal resistance machinery. ₂

Plant growth-promoting (PGP) genes were reported as 64.6% core, versus 45.2% for pathogenic genes (Mann-Whitney p=3.4e-125). The digest reads this as beneficial functions being under stronger purifying selection. That reading interprets conservation and does not directly measure selection. ₂

Essential-core genes refine the ‘metabolic engine’ interpretation. Relative to non-essential genes, they were enriched in Protein Metabolism (+13.7 pp, where pp means percentage points), Cofactors/Vitamins (+6.2 pp), and Cell Wall (+3.9 pp). They were depleted in Carbohydrates (-7.9 pp), Amino Acids (-5.6 pp), and Membrane Transport (-4.0 pp), which the digest describes as conditionally important rather than universally essential functions. ₂ The digest aligns this with Rosconi et al. (2022), who found that pan-genome composition influences essentiality in S. pneumoniae. ₂

Within a single species, gene-content variation is substantial. Across 15 PROTECT Pseudomonas aeruginosa (PA) strain groups, genome size ranged from 6.12 Mb (5,668 CDS, or protein-coding sequences) to 7.35 Mb (7,026 CDS), a 24% difference in gene content. The largest strain group (725, 98 isolates) carried 1,358 extra genes compared with the smallest. ₁₁

Despite genome sizes of 6.1–7.4 Mb and 24% gene-content variation, all 15 strain groups retained the same amino acid catabolic pathways at 97%+ conservation. This supports the conserved-metabolism side of the model within one species. ₁₁ The report attributes the extra genes to conditionally expressed accessory functions such as prophages, mobile elements, resistance islands, and niche-specific adaptations, but states this as likely rather than shown. It also expects a formulation to be equally effective across PA variants; that is a prediction, not a demonstrated result. See competitive-exclusion-consortium-design. ₁₁

Metal Fitness and Coordinated Modules

Across 22 organisms and 14 metals, genes with significant fitness defects under metal stress were 87.4% core, against a 76.9% baseline (OR=2.08, p=4.3e-162). This contradicts the project’s initial hypothesis (H1a). H1a predicted accessory enrichment for toxic-metal genes, based on a prior DvH finding of 71.2% core for condition-specific heavy-metal genes. The report attributes the discrepancy to genome-wide metal fitness defects mainly reflecting core cellular processes that metals disrupt, such as cell envelope, DNA repair, and central metabolism, rather than specialized metal resistance mechanisms. ₁₂ The result is consistent with the pooled comparison above, in which both metal-specific and general sick genes exceeded the baseline. _12,2

The same report proposes a second, accessory tier of specific metal resistance. This smaller fraction would provide targeted tolerance via efflux, sequestration, or enzymatic detoxification, and would be visible in the accessory genome only after controlling for general stress response. The tier is a proposed model, not a separately tested result. It aligns with the shared-versus-specific framing in shared-stress-versus-stressor-specific-fitness. ₁₂

Genes assigned to fitness modules were 86.0% core, compared with 81.5% of all genes, a difference of +4.5 percentage points. ₁₃ This supports the coordinated-core result from the conservation-fitness synthesis above. _13,4

The same analysis found 38 module families with <50% core genes. The report describes them as co-regulated accessory gene modules conserved across organisms. It suggests they may be horizontally transferred functional units or niche-specific operons, but these are hypotheses rather than established identities. This refines the model: coordinated functional units are not confined to the core (see module-level-coinheritance). ₁₃

Accessory Contributions to Metabolic Pathways

Pathway completeness was compared using core genes only versus all genes. The largest gaps were leucine 0.146, valine 0.146, arginine 0.141, lysine 0.140, and threonine 0.140. For these top 5 accessory-dependent pathways, the report states that approximately 14% of species-level pathway completeness comes from accessory genes. Figures figures/fig3_core_accessory_pathways.png (core versus accessory pathway completeness) and figures/core_vs_all_pathway_completeness.png (core versus all pathway completeness) show the comparison. ₁₄ This refines the conserved-metabolism side of the model. At species level, part of the capability of some amino-acid biosynthesis pathways depends on accessory genes rather than residing wholly in the core (see biosynthetic-prototrophy-and-auxotrophy). ₁₄

Carbon-source utilization pathways tended to show smaller core-versus-all gaps. The report suggests carbon catabolism genes may be either universally present or universally absent rather than variably distributed, but it presents this mechanism as speculation. ₁₄

The report contrasts these variable accessory pathway contributions with the metal_fitness_atlas finding that fitness-important genes are enriched in the core genome (OR=2.08). _14,12

In a Pseudomonas carbon-ecology comparison, amino acid catabolism (arginine, histidine, serine, glutamate, etc.) and core organic acid pathways (citrate, succinate, pyruvate) remained near-universal (>99%) in both groups compared. The four named amino-acid pathways remained >99.5% complete in P. aeruginosa despite loss of plant sugars. ₁₅ This supports the within-species 97%+ conservation of amino acid catabolic pathways across PROTECT PA strain groups described above. _15,11 These are degradation (catabolic) pathways, whereas the accessory-dependent arginine and other amino-acid pathways reported by pathway_capability_dependency are biosynthesis pathways. The two results therefore address different functions, and they also use different completeness measures and taxonomic scopes. Near-universal amino-acid catabolism neither confirms nor contradicts accessory dependence in amino-acid synthesis. _15,14

Plant-Associated Functions

All 13 plant growth-promoting (PGP) genes showed significantly higher core fractions than the 46.8% genome-wide baseline (chi-square vs baseline, Benjamini-Hochberg false discovery rate (BH-FDR) q < 0.05 for all genes). Per-gene core, auxiliary, and singleton fractions were:

• pqqC: 81.5%, 7.8%, 10.7% (q = 0.0 as reported)
• pqqB: 78.1%, 8.9%, 13.0% (q = 3.0e-247)
• hcnA: 78.5%, 10.8%, 10.8% (q = 9.2e-37)
• ipdC: 76.5%, 9.7%, 13.7% (q = 3.4e-19)
• acdS: 70.4%, 13.4%, 16.2% (q = 7.9e-25)
• nifH: 63.8%, 15.1%, 21.0% (q = 4.3e-71)
• pqqD: 55.5%, 17.0%, 27.5% (q = 2.2e-86) ₁₆

Figure figures/pgp_core_accessory_pie.png shows the per-gene fractions against the 46.8% core baseline. ₁₆

The mean accessory fraction across all PGP genes was 29.7%, compared with 53.2% genome-wide. The report treats this as clearly rejecting its hypothesis (H3) of horizontal gene transfer (HGT), the movement of genes between lineages, and states that PGP genes are inherited vertically with the core genome. That rejects an accessory- or HGT-dominated model; it does not rule out every transfer event. ₁₆ This contradicts a reading of the two-speed model in which ecologically specialized functions sit mainly in the accessory fraction. It supports the digest’s PGP-versus-pathogenic contrast above. _16,2

pqqD was an outlier at 55.5% core, against 63–81% for other PGP genes, and it had the highest singleton fraction (27.5%). The report suggests pqqD occasionally spreads horizontally as a standalone gene. It describes the functional pqq operon (pqqB–pqqC as a unit) as predominantly core. ₁₆

The primary source for the digest’s PGP contrast reports beneficial (PGP) gene clusters as predominantly core (64.6% core fraction). This exceeds both the 46.8% genome-wide baseline and pathogenic gene clusters (45.2% core fraction). A Mann-Whitney U test comparing per-species core fractions gave U=83,567,419 (p=3.38e-125), with a bootstrap 95% CI for the beneficial-pathogenic difference of [0.089, 0.106]. Elsewhere the report states the same contrast as 64.6% core vs. 45.2% for pathogenic genes (p=3.4e-125). ₁₇ See dual-nature-plant-associated-bacteria. ₁₇

By marker category, the report gives core %, singleton %, and number of clusters:

• Beneficial: 64.6% core, 20.7% singleton (N=60,590)
• Colonization: 66.6% core, 20.8% singleton (N=153,092)
• Pathogenic: 45.2% core, 31.0% singleton (N=374,416)
• Genome-wide baseline: 46.8% core, 35.3% singleton ₁₇

All chi-square tests against the baseline were significant (p≈0). ₁₇

The 46.8% genome-wide baseline is computed at the gene-cluster level across all species in the pangenome. It is the fraction of all gene clusters that their respective species pangenomes classify as core, not a species-level average. Comparisons against it should account for this definition (see pangenome-core-boundary-and-clade-size-bias). ₁₇

Among specific functions, the highest core fractions were biofilm (83.3% core), IAA (indole-3-acetic acid, an auxin) biosynthesis (78.1%), nitrogen fixation (72.3%), and phosphate solubilization (70.9%). In contrast, T4SS (type IV secretion system; 48.5% singleton), coronatine toxin (45.2% singleton), and effectors (39.8% singleton) were strongly accessory- or singleton-enriched. ₁₇ The singleton-enriched T4SS signal is consistent with the type 4 secretion and plasmid functions found among essential-auxiliary genes above. _17,5

A separate comparison in the same report gave NifH a 32.4% core-genome fraction, against 21–26% for the CWDEs (cell-wall-degrading enzymes). The report calls this consistent with its H2 pattern. ₁₇

All 50 plant-enriched OGs (orthologous groups, i.e. gene families) had core gene fractions of 60.1–83.1% (baseline: 46.8%). The report reads this as reinforcing that plant-enriched functions are stably inherited in the core genome; this Phase 2 result extended the earlier finding to 50 OGs, which the report first called “novel OGs”. Its own review later flagged “Novel OGs” as overstated and reframed them as “enriched gene families”, so they should not be read as newly discovered genes. A cross-reference with MGnify rhizosphere KEGG modules confirmed that 8 modules from these OGs are present in rhizosphere species. ₁₇

The broad core-genome result contrasts with Loper et al. (2012), who found biocontrol traits in Pseudomonas to be largely accessory. The report proposes differing taxonomic scope as the explanation: its analysis spans 25,660 species, whereas Loper et al. studied 10 strains within a single species group. This explanation has not been tested, so the PGP-core result may not hold at within-species resolution. ₁₇

Defense, Mobile, and Storage Genes

Six of seven defense-system families showed highly significant enrichment in the auxiliary and singleton pangenome. The background was 46.8% core and 37.9% singleton across 132,531,501 gene clusters. Per-system core and singleton fractions (χ² p ≈ 0 for all) were:

• Retron: 8.2% core, 65.8% singleton
• Gabija: 9.1% core, 68.8% singleton
• R-M Type II: 9.4% core, 68.2% singleton
• R-M Type I: 9.8% core, 65.3% singleton
• CBASS: 11.8% core, 61.1% singleton
• BREX: 20.7% core, 55.8% singleton
• CRISPR-Cas: 27.4% core, 47.2% singleton
• DISARM: 40.6% core, 42.3% singleton (near baseline) ₁₈

The report’s table lists R-M Types I and II as separate rows. ₁₈ This supports the COG V enrichment among singleton genes at a much larger cluster scale. DISARM sitting near baseline refines it: defense systems are not uniformly accessory (see phage-defense-syndromes-and-arms-race). _18,1

SNIPE (DUF4041) clusters were 13.3% core (604 clusters), 30.7% accessory (~1,402 clusters), and 56.1% singleton (2,566 clusters). The report interprets the combined 86.7% accessory-plus-singleton fraction as strong support for defense-island mobility and states that SNIPE is typically gained or lost rather than vertically inherited. Horizontal transfer itself was not directly measured. ₁₉

Prophage modules split along the same axis. Lysis (module D, 62.6% core) and packaging (module A, 57.0% core) were dominated by core genes. Integration (module E, 59.3% singleton) and head morphogenesis (module B, 56.0% singleton) were dominated by singletons. The report reads this as consistent with the expectation that phage structural genes diversify rapidly while regulatory and lysis genes are more conserved (Hendrix et al. 2000). ₂₀ This refines the model: even mobile elements carry their own conserved and variable fractions. ₂₀

Among 311 discordant phaC-positive species, 60.1% carried phaC as accessory genome (not present in all strains). Across all phaC-carrying species the accessory rate was 32.3%. The report takes this as strong support for horizontal acquisition, on the reasoning that recently transferred genes have not yet been fixed in the core genome; this is supporting evidence, not a direct measurement of transfer (see phb-granule-ecology). ₂₁

Overall, 5,371 species carried phaC as core and 1,959 as accessory, and some species carried both core and accessory copies. The report suggests that the high accessory fraction in SAR324, Bacillota_A, and Eremiobacterota marks these lineages as active recipients of phaC via HGT. That suggestion is a hypothesis, and it implies that a storage-metabolism gene can be core in most carriers while being accessory in particular lineages. ₂₁

Evolutionary Interpretation

The report interprets horizontal gene transfer (HGT), the movement of genetic material between lineages, as the primary innovation mechanism rather than vertical inheritance. ₁ The +10.88% enrichment of COG L was presented as evidence that most genomic novelty may come from mobile elements, although this conclusion is an interpretation of functional-category distributions rather than a direct measurement of transfer events. ₁ The central discoveries digest repeats this framing. It describes core genes as an ancient, conserved metabolic engine and novel genes as recent acquisitions for ecological adaptation. It names HGT as the primary innovation mechanism and reads the +10.88% L enrichment as suggesting that most novelty comes from mobile elements. These remain broad evolutionary interpretations rather than measured transfer rates. ₂

An ecotype analysis connects its own category L differentiation to the cog_analysis novel-versus-core enrichment of L (+10.88%) and V (+2.83%). Its category L includes transposases, integrases, and mobile genetic element machinery. This is a cross-project consistency argument, not direct evidence of mechanism in the ecotype data (see horizontal-gene-transfer-driven-innovation). ₂₂

Because the reported patterns held across the analyzed bacterial phyla, the report proposes that they reflect deep evolutionary constraint. ₁ The cross-species consistency strengthens the two-speed model, while the limited taxonomic sample leaves open the possibility that additional phylum-specific patterns would emerge in a larger comparison. ₁ The conservation-fitness synthesis supports deep conservation of coordinated core functions but qualifies a simple core-versus-accessory fitness interpretation: 28,017 genes were both costly in the laboratory and conserved in the pangenome, whereas 5,526 were costly and dispensable, making the latter candidates for ongoing gene loss rather than evidence that accessory genes are systematically burdensome. ₄ The trade-off analysis refines this interpretation: 25,271 genes, or 17.8% of the genes examined, had fitness below -1 in some conditions and above 1 in others, and trade-off genes were 1.29 times more likely to be core than non-core genes (odds ratio 1.29, p=1.2e-44). ₆ This pattern suggests that conserved genes can encode functions whose costs and benefits change across conditions, rather than representing a uniformly low-cost metabolic baseline. ₆

The essential-genome result further refines the evolutionary contrast: among 17,222 ortholog families, 859 (5.0%) were universally essential, 4,799 (27.9%) were variably essential, and 11,564 (67.1%) were never essential. ₃ Thus, conservation and indispensability are concentrated in a small subset, while many conserved or accessory genes retain context-dependent roles rather than fitting a strict core/accessory functional dichotomy. ₃

All 8 predictions from the initial N. gonorrhoeae analysis were reported as confirmed in the 32-species comparison. _1,2

Acquisition-Depth Signatures of Vertical and Horizontal Inheritance

The COG analysis inferred HGT only from category distributions. Reconstructed gain events offer a second, independent view of inheritance mode, although the report classifies the acquisition-depth spectrum and its signature plane as exploratory atlas observations: descriptive patterns, not preregistered tests. _1,23 In that exploratory spectrum, the recent-to-ancient gain ratio was described as a signature of function class. CRISPR-Cas showed a 24.5× ratio, with high recent and low ancient gains, interpreted as HGT-active. Housekeeping classes sat near ~1×, with recent ≈ ancient gains, interpreted as dominant vertical inheritance. The figure figures/p4_synthesis_H3_acquisition_depth_spectrum.png shows the spectrum. ₂₃ This is descriptively consistent with the two-speed reading that defense functions turn over by transfer while housekeeping functions are inherited. It does not confirm transfer as the mechanism; the topic is developed in gene-function-acquisition-depth. ₂₃

Leaf consistency is the fraction of species in a clade that carry a given KO (KEGG Orthology group). It was computed from 28M-row species-level KO-presence data and merged to 16.7M gain events (97.8% coverage). ₂₃ Strict housekeeping controls had high medians and low recent-gain rates: RNAP (RNA polymerase) core strict 0.97, tRNA-synth strict 0.94, and ribosomal strict 0.92. Patchy classes with active recent gains, which the report labels HGT-active, had low medians: CRISPR-Cas 0.29, two-component system histidine kinases 0.13, and AMR 0.11. Leaf consistency also declined monotonically with gain depth, from recent 0.34 to older_recent 0.30, mid 0.26, older 0.23, and ancient 0.20. ₂₃ The low AMR value is consistent with the pangenome-atlas finding that AMR is enriched in the accessory genome. _23,9

Loose housekeeping definitions include paralogs and weakly-housekeeping members. These classes sat in the middle and did not separate toward the vertical-inheritance signature as cleanly as strict definitions, so the result depends on how control classes are defined. ₂₃

Housekeeping genes also carried fewer paralogs than expected. Producer z is the report’s standardized score of paralog production relative to genes at matched prevalence. Ribosomal proteins, tRNA synthetases, and RNAP core subunits all had negative producer z (−0.15 to −0.24 σ across ranks; 95% CIs entirely below zero). This contradicted the preregistered near-zero criterion, and the report attributes it to dosage constraint. ₂₃ At Phase 1B scale, these negative controls sat ~12 % below their cohorts with z ≈ −0.15. ₂₃ This refines the conserved-core side of the model: core informational genes appear constrained in copy number as well as in presence, though the dosage explanation is the report’s interpretation. ₂₃

Composite Functional Categories

Composite COG assignments containing multiple functional letters were treated as biologically meaningful rather than as annotation artifacts. ₁ The LV composite, representing mobile and defense functions, showed +0.34% enrichment with 76% consistency and was interpreted as evidence for multifunctional modules such as mobile defense islands. ₁

The analysis recommends retaining composite COG categories because they may represent genuine multifunctional genes rather than noise. ₁ Composite categories were counted once per gene rather than split across their component letters, a choice that preserves the gene-level interpretation of multifunctionality but affects how category frequencies should be compared with analyses using split assignments. ₁

Limits of the Model

COG annotations covered approximately 70% of genes, so unassigned genes may skew the observed distributions. ₁ The analysis used 32 species, and a larger sample could reveal phylum-specific patterns not visible in the current comparison. ₁ Composite categories were counted once per gene, and eggnog v6 annotations may differ from original COG assignments. ₁

The two-speed model should therefore be treated as a strongly supported cross-species pattern in the analyzed dataset, not as a complete account of all bacterial genome evolution. ₁ The essential-genome comparison has a related limitation: the 48-organism set is taxonomically limited and biased toward culturable Proteobacteria, so the apparent distribution of conserved and lineage-specific essentials may not generalize to uncultured lineages, Actinobacteria, or Firmicutes. ₃ The laboratory fitness measurements in the conservation-fitness synthesis do not directly establish fitness in natural environments, so the costly-and-conserved category is evidence for, rather than a direct measurement of, natural purifying selection. ₄ The new conservation comparison also depends on pangenome coverage and classification: 44 of 48 Fitness Browser organisms were linked, but only 33 entered downstream analysis; clades with only 2 genomes can have trivially high core fractions, and essentiality was inferred from RB-TnSeq, a random-barcode transposon sequencing method, under library-construction growth conditions. ₅ The trade-off classification also depends on the definition of burden as fitness above 1, which may capture condition-dependent trade-offs rather than true dispensability, and the Fitness Browser conditions are biased toward experimentally convenient rather than ecologically representative settings. ₆

Tensions

A prediction derived from the two-speed framing failed in dark genes. The hypothesis that stress-condition dark genes would be more accessory than carbon/nitrogen dark genes was rejected (Fisher’s exact test, n=7,491 dark genes). The result ran in the opposite direction: stress-condition genes were 23.0% accessory versus 25.5% for carbon/nitrogen genes (OR=1.15, p=0.013). Figure fig16_h1b_test.png shows the test. ₂₄ This contradicts a simple dichotomy of stress genes as accessory and metabolism genes as core, and indicates that the relation between condition specificity and pangenome conservation is more complex. ₂₄

The cog_analysis report and the discoveries digest describe the COG pattern as holding universally across bacterial phyla and as suggesting deep evolutionary constraint. The same report also states that its 32-species sample might be too small to reveal phylum-specific patterns. The universality claim is therefore an extrapolation from the sample rather than a demonstrated result. _1,2

Functional classes do not map cleanly onto the core/accessory split. COG E (amino acid metabolism) was core-enriched, reported as -1.81% enrichment with 81% consistency ₁. Yet essential-core genes were depleted in Amino Acids by -5.6 pp relative to non-essential genes ₂. Intrinsic AMR genes were core residents even though AMR overall was enriched in the accessory genome ₉. These comparisons use different category schemes (COG versus SEED) and different contrasts (core versus singleton, essential versus non-essential), so they are not direct contradictions. They do show that metabolism and defense each contain both conserved and variable components. _1,2,9

The projects report different genome-wide singleton baselines alongside the same 46.8% core baseline. plant_microbiome_ecotypes gives 35.3% singleton, while phage_defense_arsenal gives 37.9% singleton across 132,531,501 gene clusters. Neither report explains the difference, so comparisons of singleton fractions across these projects should not assume a shared baseline. _17,18

Nitrogen-fixation conservation figures also differ by project and measure. pgp_pangenome_ecology reports nifH clusters as 63.8% core ₁₆. plant_microbiome_ecotypes reports nitrogen fixation as 72.3% core in one analysis and NifH as 32.4% core-genome fraction in another ₁₇. The sources do not reconcile these values, so the degree of nitrogen-fixation core conservation remains unresolved. _16,17

Open Directions

• Add additional taxonomic groups and repeat the COG comparison to test whether the reported +10.88%, +2.83%, +1.64%, -4.65%, -2.09%, -2.06%, -1.81%, and -1.75% contrasts remain consistent across a broader phylogenetic sample. ₁
• Examine COG L and COG V genes with gene-neighborhood, mobility, and defense-system analyses to test whether their enrichment reflects transferred elements, independently evolving defense loci, or both. ₁
• Join novel-gene COG assignments to environmental metadata and use habitat-stratified comparisons to test whether accessory functions vary by environment. ₁
• Reanalyze unassigned genes with complementary functional annotation and compare the result with the approximately 70% COG-annotated fraction to determine how annotation gaps affect the two-speed pattern. ₁
• Compare gene-level counting of composite COG assignments with component-split counting to test how representation choices affect the reported LV and other composite-category signals. ₁
• Map the 28,017 costly-and-conserved and 5,526 costly-and-dispensable genes onto COG categories and coordinated modules, then test natural-environment fitness or longitudinal gene-loss signals to distinguish retained core functions from accessory genes undergoing loss. ₄
• Stratify the 25,271 trade-off genes by COG category and assay condition, then test whether the category-specific burden reversals persist after matching organisms and environmental conditions. ₆
• Test the 7,084 orphan essentials with broader homology, mobile-element, and lineage-specificity analyses, then compare their COG and module profiles with conserved essentials to determine whether accessory indispensability reflects innovation, annotation gaps, or undetected homology. ₃
• Reclassify essential, auxiliary, and unmapped genes across matched pangenome clades using condition-specific fitness and broader homology searches to test whether the 1.56 median odds ratio persists beyond RB-TnSeq library conditions and whether poorly annotated essential genes are truly lineage-specific. ₅
• Run the leaf-consistency and recent-to-ancient gain analysis as a preregistered test on COG L, V, and S genes from the 32-species comparison. This would test whether the +10.88%, +2.83%, and +1.64% singleton enrichments correspond to HGT-active acquisition depth rather than only to category frequency, and would move the exploratory spectrum toward confirmatory evidence. _1,23
• Repeat the H1b stress versus carbon/nitrogen accessory test on all accessory genes, stratified by COG category, rather than only the 7,491 dark genes. This would show whether the opposite-direction result is specific to poorly annotated genes. ₂₄
• Map the 1,330 Acinetobacter reactions back to gene clusters, with and without gapfilled reactions. This would test whether the 94% reaction-level core persists at the gene level. ₈
• Test the predicted equal formulation response across the 15 PROTECT PA strain groups with strain-resolved growth assays. In parallel, annotate the 1,358 extra genes of group 725 for prophage, mobile-element, and resistance-island content to check whether they are accessory functions as presumed. ₁₁
• Recompute the nifH/NifH core fractions from pgp_pangenome_ecology and plant_microbiome_ecotypes with a shared cluster definition and baseline. This would show whether the 63.8%, 72.3%, and 32.4% figures differ because of measure or because of gene set. _16,17
• Test whether the 38 module families with <50% core genes co-occur with mobile-element or defense-island markers. This would check the report’s horizontal-transfer hypothesis for these accessory modules. ₁₃
• Separate metal-important genes into the proposed general-stress and specific-resistance tiers using efflux, sequestration, and detoxification annotations. Then compare the core fraction of each tier separately, which would test whether the specific tier is accessory-enriched once the general stress response is controlled. ₁₂
• Compare amino-acid biosynthesis and amino-acid catabolism completeness within the same species and with the same core-versus-all method. This would test whether accessory dependence is specific to synthesis or reflects measure differences between pathway_capability_dependency and pseudomonas_carbon_ecology. _14,15
• Repeat the PGP core-fraction comparison within single species groups such as Pseudomonas. This would test whether the cross-species core result conflicts with the strain-level accessory pattern reported by Loper et al. (2012). ₁₇