Review papers with raw data transparency

Concise critical summary

The authors analyze 7 years of monthly coastal metagenomes from BBMO and SOLA to test whether functional redundancy drives emergent, rhythmic, and synchronous metabolic dynamics; they report strong functionlevel rhythmicity and examples where functions remain rhythmic and synchronous despite high taxonomic turnover, supporting an emergent selforganization model driven by functional redundancy Paper DOI 10.1101/2025.08.12.669827

Full review and critique

1. What the paper did

The study used monthly shotgun metagenomes from two Mediterranean coastal time series, BBMO and SOLA, sampled January 2009 December 2015 (84 and 90 samples respectively) to compare synchrony and rhythmicity across three organizational levels functions, predicted genes (ORFs), and taxa (OTUs). They annotated ORFs against eggNOG KEGG Pfam COG CAZy and used LombScargle periodograms (PNmax threshold >= 8) for rhythmicity and codyn methods for synchrony analyses, focusing additionally on a curated set of 45 key biogeochemical functions to illustrate contrasting scenarios of functional redundancy Methods summary

2. Key results

• Functionlevel signals were on average more rhythmic than gene or taxon level signals particularly at SOLA where functional rhythmicity exceeded gene and OTU rhythmicity SOLA functional rhythmicity
• Average synchrony across sites was low across levels but many individual functions (including key C N P S functions) exhibited high synchrony and rhythmicity despite low synchrony/rhythmicity of their dominant genes, which the authors interpret as evidence for functional redundancy enabling emergent dynamics (examples: narH rhythmic PNmax 13.37 high turnover mean turnover 0.82; amoB PNmax 22.7 dominated by Nitrososphaeria with low turnover) narH and amoB examples
• Comparisons with global Tara Oceans catalogs showed taxonomic composition per function broadly coherent with global patterns for many functions (Pearson mean ~0.62 though only 12 functions reached statistical significance) suggesting partial generality beyond the two sites Comparison with Tara

3. Strengths

• Long highfrequency time series across 7 years with consistent monthly sampling provides statistical power to detect seasonality and temporal structure in a marine context Sampling depth
• Multilevel analysis (functions genes taxa) combined with turnover and dominance metrics and formal synchrony and rhythmicity statistics gives a mechanisticfriendly framing linking redundancy to emergent dynamics rather than simple correlation.
• Transparent identification of potential confounders and choices (PNmax threshold dominance cutoffs occurrence filters) and several sensitivityaware analyses for illustrative functions.

4. Limitations and potential biases

1. Annotation and inference from short reads Functional assignments are drawn from shortread assemblies and database mappings (eggNOG KEGG Pfam etc) which can misannotate multifunctional or paralogous genes and create false positives especially for enzymes with conserved domains (authors note coxL CODH cleaning via phylogenetic placement). This limits confidence that ORF counts equal realized metabolic rates coxL cleaning
2. Function abundance versus activity Metagenomic ORF abundance does not equal transcriptional or enzymatic activity; functions may be present but not expressed, or activity may be regulated posttranscriptionally. A complementary metatranscriptome or rate measurement would be needed to confirm that emergent rhythmicity in function abundance corresponds to fluxes.
3. Threshold sensitivity and filtering choices Key numerical thresholds shape which features are labeled rhythmic or dominant (PNmax >= 8 dominant gene cutoff 70% occurrence cutoff 30%). Results could shift if thresholds change; authors acknowledge this but a systematic sensitivity sweep is not fully shown Threshold limitations
4. Crosssite environmental heterogeneity Although the sites are connected and broadly similar, SOLA shows higher environmental variability (river inputs Tramontana winds) which could drive higher functional rhythmicity independently of redundancy; disentangling environmental forcing versus internal redundancydriven selforganization requires causal or perturbation data.
5. Data availability and reproducibility The paper states no data availability location; raw read and processed tables appear not to be linked which reduces reproducibility; given large datasets (~41 billion reads) deposition of raw reads and processed feature tables is essential for independent validation Data availability absent

5. Interpretation and alternative explanations

The central claim that functional redundancy enables emergent rhythmic and synchronous functionlevel dynamics is consistent with the observations that functions can show high rhythmicity while their dominant genes are nonrhythmic and turnover is high. This is a plausible mechanistic interpretation because multiple taxa/timevarying contributors can sum to a stable functionlevel timecourse. However alternatives exist:

• Environmental synchronization of different taxa via common external drivers (light temperature nutrient pulses) could cause convergent timing of function realizations even without compensatory redundancy; showing that functionlevel rhythms persist after statistically removing measured environmental covariates would strengthen the causal claim.
• Populationlevel dormancy or activity cycles that affect many taxa similarly could produce synchrony at function level via synchronous regulation not just taxon replacement.

6. Recommendations to strengthen claims

1. Publicly release raw reads and processed abundance tables per sample with annotation mappings to eggNOG KEGG Pfam ORFtofunction maps and OTU tables to permit reproduction and reanalysis.
2. Include metatranscriptomic or targeted rate measurements for a subset of functions (e.g., ammonia oxidation nitrate reduction CO oxidation) to show that function abundance rhythms correspond to activity rhythms.
3. Perform sensitivity analyses varying PNmax thresholds dominant gene cutoffs and occurrence filters and report how many functions change classification under alternate thresholds.
4. Using causal models (e.g., partial canonical correspondence analysis time series Granger causality with environmental covariates) to partition environmental forcing versus internal compensatory dynamics.

7. Bottomline evaluation

The paper presents a careful, wellpowered temporal metagenomic analysis and advances the argument that functional redundancy can generate emergent rhythmic and synchronous metabolic dynamics in marine microbiomes. The evidence is persuasive for the existence of functionlevel rhythmicity and instances where taxonlevel turnover supports stable function timing. However inferential leaps linking abundance to activity and the absence of public data lower the reproducibility and causal strength of the claims. Overall the work is a significant contribution that motivates followup work combining metatranscriptomes rates and perturbations.