Review Papers Against Their Own Results

 Quick Explanation Copied

 Long Explanation

What the paper does well

The review accurately traces the field from the FANTOM-imposed 300-nt annotation cut-off through RIBO-Seq, MS-based microproteomics, proteogenomics, and CRISPR/coIP functional validation, correctly emphasizing that ribosome occupancy alone is not proof of coding function—a point originally established by Guttman et al. The reviewed paper itself. Its tool and database tables (sORFfinder, MiPepid, sORFs.org, SmProt, OpenProt, MetamORF) are useful entry points MiPepid benchmark metrics. The authors fairly note MS under-detection biases: bottom-up proteomics favors >10 kDa proteins (over 90% of annotated proteome) and single-peptide matches inflate FDR for tiny ORFs.

Figure 1: This funnel (log scale) uses reported values from the mouse genome-wide study the review cites as an exemplar of prediction-plus-Ribo-Seq integration: 40.7M predicted sORFs narrowed to 2.4M high-coding-potential candidates (SVM 93% accuracy, 4% FPR), then to a few hundred with ribosome evidence Mouse sORF prediction + Ribo-Seq study. It visually exposes the field's central problem: the candidate-to-function gap of ~4 orders of magnitude.

Where it falls short

• Redundancy, not novelty: the field-leading Schlesinger & Elsässer review already covered Ribo-seq, MS, proteogenomics, and validation (including APEX tagging, split-fluorescent tagging, frameshift rescues) a year earlier, with 202 references Comprehensive earlier sORF review. The abstract's claim that its "novelty lies within its validation for the functional role of microproteins" is not substantiated by any benchmark, experiment, or decision framework.
• No prioritization logic: the review says candidates should be "narrowed down" (e.g., differential expression in K562/MOLT4) but offers no quantitative framework for triaging millions of candidates—a gap now partially filled by database-scale resources Bacterial sORFdb scale.
• Annotated-text errors: the figure legend describing sORF localities states "3' UTR (uORF), 5' UTR (dORF)"—reversed terminology (uORF = upstream/5'). Such slips, plus garbled reference metadata, reduce trust.
• Bacterial omission: the review frames sORFs mainly as a eukaryotic/lncRNA story, yet bacterial small-protein discovery (Ribo-seq found 338 unannotated sORFs in uropathogenic E. coli CFT073, with MS detecting only ~7—illustrating the MS sensitivity gap the review describes) is a major parallel literature UPEC small-protein census.

Confidence and what would change the verdict

Confidence in this assessment is moderate-to-high: it rests on the paper's own text and directly comparable reviews. The verdict would change if the review's tool tables were shown to be materially more current or comprehensive than the 2021 FEBS review for a specific user need. Hidden assumption to watch: the review treats multi-omics concordance as incrementally validating microproteins, but overlapping peptides and shared homologous regions can make even proteogenomic hits ambiguous—its own citation of inflated search-space false positives acknowledges this without resolving it Proteogenomics FDR caveat.

 BGPT Paper Review

Study Novelty

An incremental update overlapping heavily with the 2021 Schlesinger & Elsässer FEBS review; its claimed novelty (functional validation focus) is partially duplicated by that earlier work.

Scientific Quality

Well-organized and honest about method limitations, but contains terminology errors (uORF/dORF reversal in figure legend), garbled reference metadata, no quantitative benchmark of its own, and no original synthesis beyond prior reviews.

Study Generality

Covers tools, databases, and methods spanning human, mouse, plants, Drosophila, yeast, and mitochondria, though it omits the substantial bacterial small-protein literature.

Study Usefulness

Its compiled tool/database tables and worked functional-validation examples (Mtln, SPAR, MLN, humanin) are practical entry points for researchers designing sORF studies.

Study Reproducibility

Review articles cite methods rather than implement them; tool and database URLs are listed but without versioning, benchmark datasets, or decision workflows that a reader could directly execute.

Explanatory Depth

Describes mechanisms of RIBO-Seq, MS biases, and proteogenomic FDR issues at a competent review level but does not derive or test causal explanations for why false positives persist.