Review papers with raw data transparency

Critical analysis — claims, evidence, and limitations

1) Do the data support proteome expansion driven by HGT in industrialized hosts?
Evidence: reported per-species increases in median CDS counts for industrialized strains, consistent higher pangenome fluidity across species, reconciliation showing elevated gene-gain counts on terminal branches associated with industrialized-host tips, and correlation between gain–loss differences and per-branch HGT counts — all consistent with recent HGT-mediated acquisition of accessory genes rather than ancient divergence GMbC isolate + reconciliation evidence.

Assessment: The multi-pronged approach (isolates, reconciliation, HGT BLAST pipeline) provides convergent evidence that HGT contributed to proteome enlargement in industrialized-associated strains. However, distinguishing adaptive acquisition from accumulation due to relaxed purifying selection requires (and the authors acknowledge) demographic/Ne estimates, temporal sampling, or fitness assays. The paper correctly notes this ambiguity in the Discussion.

2) Are MAGs failing to capture accessory/MGE content?
Evidence: direct paired comparisons (n=147) show isolates are larger (median +~458 kb) and contain more CDS and MGEs; isolates detect more HGT events (species-pair proportion significantly higher; X2=142.4, p=7.96e-33). This aligns with prior critiques that MAGs under-represent mobilome and accessory elements due to assembly/binning limits Isolate vs MAG paired comparison and consistent with prior reports about MAG limitations Shaiber & Eren (mBio).

Assessment: The paired analysis is a strong methodological contribution. It justifies privileging isolates for accessory/mobilome analyses and using MAGs only for validation and broader sampling. Nevertheless, the study's MAG set is short-read; long-read metagenomics (HiFi/ONT) would be needed to fully close the MAG–isolate gap (authors note this).

3) Convergence across species: are gene- and SNV-level signals convincing?
Evidence: Gene presence/absence regressions (phylogeny-aware) identified ~5% of gene families per species associated with lifestyle; six gene families recur across ≥2 species (e.g., ugd, wecA, cas1, traM) and multiple tra/relaxase elements appear in cross-species clusters, indicating mobilome-mediated convergence. Ka/Ks and SNV analyses (with recombination masking) identify lifestyle-specific positive selection and high-confidence non-synonymous SNVs validated in metagenomes (GMbC n=1,015). The cob operon and mur/lpx/rfb operons recur across species as SNV targets — functional plausibility exists (vitamin B12 biosynthesis, peptidoglycan/LPS remodeling) Cross-species gene & SNV convergence.

Assessment: The convergence claims are well supported by (i) cross-species replication, (ii) functional coherence (cell envelope, stress response, metabolism), and (iii) validation in metagenomes. Limitations: many cross-species hits are unannotated; statistical multiple-testing and population structure remain concerns despite phylogeny-aware models; causality (selection by a specific environmental factor) is not established — industrialization bundles diet, antibiotics, sanitation, hosts genetics, and environment, any of which could drive the signals.

Major limitations & blindspots (explicit)

• Binary industrialization label simplifies a multidimensional phenotype — authors use PC1 Lifestyle to help but residual confounding remains (diet, antibiotics, urban environment, host genetics) Poyet et al. (GMbC cohort).
• Analyses emphasize 10 species for depth; generality across the wider gut microbiome remains to be shown (authors partially validate with MAGs but MAGs undersample MGEs).
• Isolate culture bias: cultivated strains may not represent in situ strain diversity; culturing selects for certain physiologies (a known limitation; authors acknowledge).
• Proteome expansion could reflect (A) adaptive gene acquisition via HGT, (B) relaxation of purifying selection and accumulation of MGEs, or (C) sampling/culturing artifacts. Distinguishing these requires Ne estimates, temporal samples, or fitness measures (authors note this).
• MAGs derived from short-read assemblies limit resolution of plasmids and integrative elements; long-read metagenomes would strengthen claims about mobilome dynamics.

Practical recommendations & next experiments

1. Functional tests: clone candidate LPS/ugd/wecA/cob variants and measure phenotypes (e.g., LPS structure, cobalamin production, cell envelope resistance to bile/oxidative stress) in representative strains to test adaptive benefit.
2. Fitness and competition assays: in vitro gut-simulators (anaerobic chemostats) or gnotobiotic mice colonized with paired industrial/non-industrial alleles to measure fitness under industrialized diet/antibiotic regimes.
3. Long-read metagenomics (PacBio HiFi/ONT) on samples with cultured isolates to resolve plasmids/prophages and validate HGT trajectories and element structures.
4. Population-genetic inference: estimate effective population sizes, recombination rates, and selection coefficients per species to partition adaptive vs nonadaptive contributions to proteome expansion.
5. Temporal sampling: longitudinal sampling across lifestyle transitions (e.g., urbanizing communities) would test the timing of gene gains and allele frequency shifts.

Confidence, final appraisal

Overall this is a high-quality, carefully controlled genomics study that advances our understanding of how host lifestyle correlates with rapid genomic change in gut commensals. The isolate resource and paired analyses are particular strengths. Main open questions are mechanistic: which environmental factor(s) (dietary substrates, antibiotics, sanitation-driven community composition) select for the observed gene gains, and to what extent expansion is adaptive vs tolerated genetic baggage. The authors are appropriately cautious in their interpretation and suggest follow-up experiments.

Selected citations used in this review (key sources)