Review Papers Against Their Own Results

Overview of study and main claims

• The authors performed a comparative genomic and transcriptomic analysis across Poaceae and outgroups to map presence/absence, phylogeny, synteny, and coexpression of anthocyanin and proanthocyanidin pathway genes, concluding an evolutionary rewiring in grasses led by a Poaceae-specific MYB (ncaMYB) and a Poaceae-specific arGST (BZ2), with complex losses (eg ANS absence) in Brachypodium Poaceae Anthocyanin Rewiring Preprint

Methods summary (what they did)

• Genome collection and annotation: used public Poaceae genomes plus re-predicted CDS for 13 unannotated genomes using GeMoMa (homology-based gene prediction) and applied BUSCO for completeness checks Data collection and BUSCO checks
• Gene detection and filtering: KIPEs v3 for pathway enzyme baits; MMSeqs2 and curated bait sets; filtering by conserved residues and curated pipelines available on GitHub (Poaceae_ABP) Gene detection with KIPEs
• Phylogenetics: alignments (MAFFT, Muscle), trimmed (trimAl), and IQ-TREE ML trees with ModelFinder and 1000 bootstraps to infer independent lineages and duplication/loss events Phylogenetic pipeline
• Expression/coexpression: large public RNA-seq sets quantified with kallisto and analyzed with CoExp to support functional assignment (coexpression of candidate arGSTs and MYBs with structural genes) RNAseq quantification and coexpression
• Synteny: JCVI/MCScan used to test microsynteny around key loci (MYB C1/Pl and ANS) to infer gene loss versus relocation Synteny analysis

Key results with appraisal

1. Independent anthocyanin activator MYB lineage (ncaMYB): Authors show Poaceae MYB sequences cluster separately from canonical caMYBs and instead group with maize C1/Pl and rice OsC1-like sequences, defining a Poaceae-specific clade named ncaMYB. Phylogenetic support and coexpression evidence are presented to argue regulatory takeover by ncaMYBs in grasses ncaMYB phylogenetic evidence Appraisal: the phylogenetic clustering is convincing for a distinct lineage; however, sequence-based phylogeny alone cannot prove functional equivalence to canonical activators without biochemical or transgenic functional assays. The authors partially mitigate this by coexpression analyses, but direct functional validation (e.g., heterologous expression, complementation assays) is missing and is a key next step.
2. Independent arGST lineage (BZ2): Two independent arGST lineages are reported: Poaceae BZ2 tau-GST clade is non-orthologous to the TT19/An9 lineage found in many angiosperms; coexpression with pathway genes supports an anthocyanin-related role arGST BZ2 lineage Appraisal: the evidence for independent recruitment is plausible; experimental demonstration of anthocyanin-binding or transport activity (biochemical assays, complementation of arGST mutants) would strengthen the claim beyond phylogeny+coexpression.
3. Proanthocyanidin pathway remodeling: ANR shows two Poaceae-specific clades (ANRPo1/ANRPo2) with distinct coexpression patterns; LAR is missing in multiple Panicoideae tribes suggesting repeated losses Proanthocyanidin dynamics Appraisal: robust phylogenetic sampling is needed to time these losses; authors provide good taxon coverage but the pattern requires denser sampling in undersampled clades to exclude assembly gaps or annotation misses.
4. Loss of anthocyanin biosynthesis in Brachypodium: The study reports absence of ncaMYB, BZ2, and ANS orthologs in Brachypodium spp., inferring pathway loss in that genus Anthocyanin loss in Brachypodium Appraisal: this is a striking claim with plausible support, but it is vulnerable to false negatives from assembly gaps or divergent sequences; authors correctly used TBLASTN checks and synteny to reduce false negatives, but wet-lab pigment assays or metabolomics in Brachypodium would be decisive.

Strengths

• Comprehensive multi-tool pipeline combining gene prediction (GeMoMa), targeted gene discovery (KIPEs), rigorous phylogenetics (IQ-TREE with ModelFinder and high bootstraps), synteny (MCScan), and expression/coexpression analyses (kallisto, CoExp) — scripts and pipelines are publicly available (GitHub) increasing reproducibility Data and scripts availability
• Convergent lines of evidence (phylogeny, coexpression, microsynteny) are used to support evolutionary claims rather than a single method.

Limitations and potential biases

• Functional claims rely primarily on in silico evidence (phylogeny, coexpression). Functional biochemical validation (e.g., complementation assays, binding/transport assays for GSTs, transactivation assays for MYBs) is lacking and necessary to confirm mechanistic roles.
• Presence/absence calls are sensitive to genome assembly and annotation completeness. Although the authors re-predicted genes and used TBLASTN checks, some reported absences (notably ANS in Brachypodium) could still reflect assembly or annotation gaps or high divergence; metabolite profiling in Brachypodium would validate loss claims Checks for false negatives
• Taxon sampling matters: Poaceae is very large (~12,000 species) and while the study covers many important clades, denser sampling in key lineages (especially early-diverging grasses and lesser-studied PACMAD/BOP subclades) would refine timing of duplications/losses.
• Coexpression evidence is correlative and can be confounded by tissue sampling biases and experimental conditions across heterogeneous SRA datasets; the authors use coexpression as supporting evidence, which is appropriate but not definitive for function.

Reproducibility and robustness

• Code availability: primary pipelines and helper scripts are available at https://github.com/ajones292/Poaceae_ABP and related repositories for expression and phylogeny workflows (CoExp, PlantGenomicsGuide) which facilitates reproduction Poaceae_ABP GitHub
• Data availability: authors indicate all underlying datasets are public (SRA) and list Additional Files, enabling reanalysis by others.
• Methodological transparency: detailed methods and tool versions are listed, aiding exact reproduction.

What evidence would falsify the main conclusions

• Experimental demonstration that canonical caMYBs (or TT19/An9 orthologs) are present and functionally active in Poaceae species previously assigned to ncaMYB/BZ2 regimes would contradict the independent evolution model.
• Discovery of ANS, ncaMYB, or BZ2 orthologs in Brachypodium through long-read assemblies or targeted PCR/metabolite detection of cyanidin derivatives would falsify the inferred pathway loss in that genus.
• Biochemical assays showing that BZ2-type GSTs do not bind/transport anthocyanins would undermine the proposed functional replacement of TT19/An9.

Recommended next experiments (to strengthen claims)

1. Functional validation: Heterologous complementation of Arabidopsis tt19 or ANS mutants with Poaceae BZ2 and ncaMYB genes, respectively, to test rescue of pigmentation phenotypes.
2. Biochemical assays: in vitro binding or transport assays for BZ2 proteins (ligandin assays, GST enzyme assays) and electrophoretic mobility shift or transactivation assays for ncaMYB binding to promoters of ANS/DFR/UFGT.
3. Targeted metabolomics: LC-MS profiling of anthocyanins and proanthocyanidins in Brachypodium accessions to confirm biochemical absence/presence correlated with genomic predictions.
4. Broader taxon sampling: add early-diverging Poaceae and undersampled PACMAD/BOP members with high-quality assemblies to refine timing of duplications and losses.

Final evaluation (balanced)

The paper presents a coherent and well-documented in silico case that anthocyanin biosynthesis regulation in Poaceae has been remodeled via independent lineage evolution (ncaMYB, BZ2) and lineage-specific structural gene losses (ANS, LAR patterns), with a provocative claim that Brachypodium has lost anthocyanin biosynthesis. The combination of phylogenetics, synteny, and coexpression provides substantial support; however, the major limitation is the absence of direct functional or biochemical validation. With targeted wet-lab confirmation (complementation, metabolomics, binding assays), the conclusions would become much stronger and mechanistically convincing.

Actionable reproducible resources: GitHub repo with scripts: https://github.com/ajones292/Poaceae_ABP (pipeline for gene detection, phylogeny, presence-absence matrices) — re-running the published pipelines on public SRA data should reproduce the main in silico findings Poaceae_ABP repo description