A conserved glycan motif induces broadly reactive functional antibodies against the zoonotic pathogen Streptococcus suis.
The main results reproduced: recomputed values matched the published ones within tolerance.
Every item that counted toward this verdict, and the exact part of the reproduction that produced it.
- ✓The central claim held under reproduction
- 🟡Could not use the authors’ exact input data
- 🟡Reported values were only indirectly comparable
- 🟡A deviation arose in the data or preprocessing
- 🟡A deviation was attributed to the published material
- 🟡Reported values were not (fully) derivable from the shared data
- 🟡The deviation was non-trivial in magnitude
- 🟡Overall, the reproduction showed a material discrepancy
A 0–100 reproducibility-quality score from the per-question grades, shown as a z-score: standard deviations above (+) or below (−) the mean of comparable assessments.
▸Reproduction agent’s raw note
P16 third-party-tool reproduction, run fully from scratch (prior «infra» workdir was reclaimed by the janitor). This is a glycobiology/immunology paper; the only pipeline-derived, code-linked result is the ABRicate (github.com/tseemann/abricate) conservation screen of the 5 dTDP-rhamnose + 14 RPS biosynthesis genes of strain P1/7 (AM946016.1 = NC_012925.1) across a public S. suis genome collection (conservation index = %coverage x %identity), grouped by MLST sequence type. ENTIRE pipeline executed inside ONE self-contained SLURM job on a compute node (env-build + genome download + QC + ABRicate + MLST + aggregate; nothing on front1): «job», COMPLETED 00:37:10 on node n117. I reconstructed the 19-gene query DB from P1/7 (19/19, Table-S1 SSU#### locus tags), re-collected the 2026 GenBank S. suis snapshot (8860 -> 8830 rehydrated; 30 assemblies withdrawn/suppressed at NCBI), and applied the paper's FULL 5-criteria QC (contigs<=500, N50>=10kb, 1.6-3.0 Mbp, GC 40.0-42.5%, uncalled-N<=50) -> 7380 pass (note: this corrects a prior run that had applied only the first 3 criteria). RESULT: the paper's substantive scientific claims REPRODUCE. C2 (rhamnose genes ubiquitous + conserved) = exact (rml genes in 99.7-100% at 96.6-99.6% conservation). C3 (ST-1/ST-7 all RPS >91.5%) = exact (every RPS gene median = 100.0 in both STs; zero genes below threshold). C4 (ST-16/ST-20 srpC & srpL <86.5%) = within-tol: srpL median 84.64 (<86.5) in both STs and is the single most divergent RPS gene overall; srpC median 86.56 sits 0.06pp above the stated 86.5 threshold (snapshot/rounding level) but srpC & srpL are correctly THE divergent pair -- the paper's central genomic finding that SrpC/SrpL glycosyltransferases drive RPS allelic diversity. C4b also holds (ST-25 srpL 84.44, ST-28 srpL 84.54 both divergent). NOTABLY, the per-ST C4 medians (86.56 / 84.64) are IDENTICAL to an independent earlier run on a different QC-filtered set -- the conservation values are deterministic and robust to the exact genome set. The count claims C1 (1719) and C5 (1350) = partial: NOT byte-reproducible because the paper's exact genome accession list is deposited NOWHERE (absent from the supplement, from Zenodo 10.5281/zenodo.18218680 which holds only a 21-kB helper script, and the data-availability statement is generic boilerplate); the 2026 snapshot is ~5x larger. NO fabrication signal -- the undeposited list is a reproducibility-transparency gap, and every graded value is independently derivable. NOT ATTEMPTED: all wet-lab results (lectin staining, GC-MS, NMR, mutant phenotypes, ELISA/opsonophagocytosis, vaccination - out of computational scope) and the optional core-genome ML tree (no numeric claim to grade).
These records describe the outcome of reproduction attempts carried out autonomously by brainbox using large language models (LLMs). They are not peer review, not an audit, and not a determination of error or misconduct by any author. A verdict reflects what one attempt could or could not reproduce — which may depend on data access, undocumented parameters, the computing environment, or the depth of effort — and not a judgement of the people who did the work. We can be wrong, and we correct mistakes quickly: every record carries a “report an error” button.
Assessment versions
Every reproduction run is kept as an immutable version — anchored to the data as it stood, with a tamper-evident chain hash. A rerun (e.g. after an author updates a deposit) adds a new version; the previous one stays on record.
-
v1 current initial assessment Score 77assessed: 2026-06-21 ⛓ d551f403d944
✎ I am an author of this paper
Updated or fixed a deposit, or is there an erratum? Ask us to re-run the metrics. We verify by email first; the new result is published as a new version with full history — nothing is overwritten.
Provenance — full disclosure
When this reproduction was carried out, which methodology version was used, and by whom — so the record can be audited and checked independently.
- Reproduced
- 2026-06-23
- Rubric version
- v1.0
- Assessed by
-
🤖 AI curator · claude (ai-curator room) · v1.0 · run #1 2026-06-21no human curator yet
- Last updated
- 2026-08-05
Provisional, curator- or AI-assessed, and independently checkable. A reproduction outcome states what one attempt could reproduce — not a judgement of the authors.
Deep full-text extraction
Model: sonnetSimilar to Streptococcus pyogenes and Streptococcus mutans, the authors hypothesized that Streptococcus suis expresses rhamnose-rich cell wall polysaccharides (RPS) with limited structural variation compared to capsular polysaccharide (CPS) variation, and investigated whether a conserved RPS glycan motif could serve as a broadly protective vaccine target.
- ★ Pathogenic S. suis lineages express two structural RPS variants that differ by presence/absence of glucose but share a conserved glycan core finding
- ★ srpC and srpL glycosyltransferase genes show allelic diversity and underlie RPS structural variation between pathogenic S. suis lineages finding
- srpC and srpO are essential for S. suis viability, as deletion mutants could not be obtained despite multiple attempts finding
- ★ Deletion of srpP removes Gal/Glc side chains and reveals an identical conserved RPS glycan core of Rha and GlcNAc in both strains mechanism
- ★ Immunization of pigs with the conserved RPS core domain elicited antibodies recognizing antigenically diverse pathogenic S. suis strains finding
- ★ RPS core-elicited antibodies induced complement deposition on encapsulated pathogenic S. suis strains, showing RPS is accessible despite capsule finding
- The S. suis RPS gene cluster (srpBCDEGIJKLMNPQR) lacks dTDP-rhamnose biosynthesis genes, which are instead encoded by a separate upstream five-gene rmlA-D cluster finding
- The conserved ΔsrpP RPS core is a linear polysaccharide with a trisaccharide repeating unit of α-configured sugar residues finding
| Assay | System | Perturbation | Readout | Platform |
|---|---|---|---|---|
| Comparative genomics (BLAST/ABRicate gene conservation analysis) | 1719 publicly available S. suis genomes | none | conservation index (identity x coverage) of RPS and rhamnose biosynthesis genes | ABRicate/BLAST |
| Plant lectin staining (SBA, RCA120, sWGA) | S. suis strains S10 and 861160 (capsule-deficient, ΔCPS) | genetic deletion (ΔsrpL, ΔsrpP) and plasmid complementation | lectin binding to bacterial surface glycans | — |
| Glycosyl composition analysis (GC-MS of TMS derivatives of methyl glycosides) | Isolated RPS from S. suis S10 and 861160 (WT and mutants) | genetic deletion (ΔsrpL, ΔsrpP) | molar ratio of monosaccharides (Rha, Gal, Glc, GlcNAc) | GC-MS |
| Glycosyl linkage analysis (GC-MS of partially methylated alditol acetate derivatives) | Isolated RPS from S. suis S10 and 861160 (WT and mutants) | genetic deletion (ΔsrpL, ΔsrpP) | identity of glycosyl linkage residues | GC-MS |
| Phosphate quantification (malachite green assay) | Isolated S. suis RPS from S10 and 861160 | none | presence/absence of phosphate | malachite green assay |
| NMR spectroscopy (1H,13C-HSQC and F2-coupled HSQC) | Isolated ΔsrpP RPS from S. suis S10 and 861160 | genetic deletion (ΔsrpP) | glycan structure, anomeric configuration, coupling constants | NMR |
| Size exclusion and DEAE ion-exchange chromatography | S. suis S10ΔCPS cell wall extract | mild acid hydrolysis after chemical N-acetylation | purification/fractionation of charged vs uncharged RPS | DEAE chromatography |
| Homologous recombination gene deletion and heterologous complementation | S. suis strains S10 and 861160 | KO (srpC, srpL, srpP, srpO) and complementation (psrpL-s, psrpL-c) | viability and RPS phenotype | — |
- – S10ΔCPS bound SBA and RCA120; 861160ΔCPS bound only RCA120, indicating different RPS composition between strains
- – In pathogenic ST-1/ST-7 lineages, all RPS genes showed high conservation to P1/7; ST-16/ST-20 lineages showed low conservation in srpC and srpL >91.5% vs <86.5%
- – ΔsrpL reduced SBA/RCA120 binding in S10 but increased binding in 861160; restored by homologous but not cross-complementation
- ▼ ΔsrpP abolished RCA120 and SBA binding but conferred sWGA binding in both strains, and RPS lacked Gal/Glc, containing only Rha and GlcNAc
- – S10 WT RPS contained Rha, Gal, GlcNAc; 861160 WT RPS additionally contained Glc molar ratio 5.2:2.0:2.7 (S10) vs 6.5:1.0:0.7:1.8 (861160)
- – Anomeric one-bond coupling constants indicated alpha-configuration of the three sugar residues in the conserved ΔsrpP RPS trisaccharide repeat 1JC1,H1 = 171 to 177 Hz
- ▲ Immunization with the RPS core domain elicited antibodies recognizing antigenically diverse pathogenic S. suis strains
- ▲ Core-domain-elicited antibodies induced complement deposition on encapsulated pathogenic S. suis strains
- count 1719 S. suis genomes analyzed (gene conservation analysis by ABRicate/BLAST)
- other >91.5% conservation (RPS genes in ST-1/ST-7 lineages vs P1/7 reference)
- other <86.5% conservation (srpC and srpL genes in ST-16/ST-20 lineages vs P1/7)
- other molar ratio Rha:Gal:GlcNAc = 5.2:2.0:2.7 (S10 WT RPS glycosyl composition)
- other molar ratio Rha:Glc:Gal:GlcNAc = 6.5:1.0:0.7:1.8 (861160 WT RPS glycosyl composition)
- other 1JC1,H1 = 171 to 177 Hz (NMR coupling constants confirming alpha-anomeric configuration of ΔsrpP RPS sugars)
- count 14-gene RPS biosynthesis cluster (srpBCDEGIJKLMNPQR) (S. suis RPS gene cluster composition)
- count 5-gene dTDP-rhamnose biosynthesis cluster (upstream rmlA-D cluster in S. suis P1/7 genome)
Statistical methods review
Model: sonnetA neutral, descriptive read of the statistical approach — what was done, and (for shared learning, not as criticism) what could also have been done.
This biochemistry and microbiology paper characterizes two cell wall polysaccharide (RPS) structural variants in pathogenic S. suis strains using genetic deletion, heterologous complementation, lectin-binding assays, GC-MS glycosyl composition/linkage analysis, and NMR spectroscopy. Quantitative data from lectin-binding experiments are reported as means ± SD from biological or technical triplicates. The provided text is truncated before the immunization/antibody functional assay results and the formal statistical methods section, so the full scope of inferential testing cannot be assessed from the supplied excerpt.
-
Lectin-binding fluorescence data from triplicates are summarized as mean ± SD with no formal inferential test reported in the visible text↳ Could also: A one-way or two-way ANOVA (or Kruskal-Wallis for non-normal data) with a post-hoc correction (e.g., Tukey HSD or Dunn's) applied across the multiple mutant/strain groups — With several deletion mutants compared to two parent strains simultaneously, a single omnibus test followed by corrected pairwise comparisons would control family-wise error rate and provide p-values and effect-size estimates (e.g., partial η²) that facilitate quantitative interpretation
-
Dispersion is reported as SD across triplicates (n = 3)↳ Could also: Report 95% confidence intervals instead of or alongside SD — At very small n (3 replicates), a 95% CI directly conveys inferential uncertainty about the mean and is often preferred in addition to SD, which describes sample spread; some journals and reporting guidelines (e.g., Nature Methods) now recommend CIs for small-n biological data
-
Genomic conservation across 1719 genomes was assessed by multiplying BLAST identity and coverage to form a continuous conservation index↳ Could also: Pangenome graph approaches (e.g., PPanGGOLiN, Roary with core/accessory statistics) or phylogenetic regression (BayesTraits) to formally test conservation vs. phylogenetic signal — A formal pangenome or phylogenetically corrected analysis would distinguish genuine conservation from phylogenetic non-independence among closely related strains, providing statistical confidence intervals on gene presence/absence rates
-
Technical triplicates from a single experiment are reported as representative of two independent experiments rather than pooling both experiments↳ Could also: Pool or summarize across both independent experiments (e.g., report grand mean ± SD across all independent replicates, or use a mixed-effects model treating experiment as a random factor) — Reporting one representative experiment can mask inter-experiment variability; a mixed-effects approach or explicit pooling across independent runs gives a more reproducible estimate of the true effect magnitude
-
Molar ratios of monosaccharides from GC-MS composition analysis are reported as descriptive point estimates (e.g., 5.2:2.0:2.7) without accompanying uncertainty↳ Could also: Report technical replicate means ± SD or CV% for the GC-MS molar ratios, or confirm by orthogonal quantification (e.g., HPAEC-PAD) — Providing variability around the compositional ratios would allow readers to judge whether differences between strains (e.g., presence vs. absence of Glc) exceed normal analytical measurement error, strengthening the interpretive basis for structural conclusions
-
Allelic diversity in srpC and srpL was characterized by defining a binary threshold (>90% coverage and >80% identity) for conservation calls across 1719 genomes↳ Could also: Apply a continuous Bayesian or distance-based clustering (e.g., PopPUNK, FastBAPSs) to group alleles, or use a phylogenetic signal test (e.g., Pagel's λ) to assess whether allele distribution tracks phylogeny — Threshold-based binary presence/absence calls can be sensitive to the chosen cutoffs; a model-based or continuous approach would make the allele-grouping decision data-driven and would quantify uncertainty in lineage assignments
What was reproduced
The exact results taken into scope, with each reported value next to the value our attempt produced.
Scope — pmid-41880495
Paper: Shi et al. 2026, Sci Adv — "A conserved glycan motif induces broadly reactive functional antibodies against the zoonotic pathogen Streptococcus suis." DOI 10.1126/sciadv.adz1854 · PMCID PMC13015895.
Code (P16, third-party tool): ABRicate — https://github.com/tseemann/abricate (v1.0.1).
Helper: extract_genes_ABRicate v0.0.2 (zenodo 10.5281/zenodo.18218680) — only a
21-kB python script that slices ABRicate hits out of assemblies; not data.
This is a glycobiology/immunology paper. The vast majority of results are wet-lab (lectin staining, GC-MS glycosyl composition/linkage, NMR, mutant phenotypes, ELISA/opsonophagocytosis, vaccination). Those are out of scope (cannot be reproduced computationally). One self-contained bioinformatic analysis is in scope and is tied to the cited code.
IN SCOPE (pipeline-derived)
The conservation screen of the RPS / dTDP-rhamnose biosynthesis genes across a S. suis genome collection by ABRicate (Fig. 1B; Methods "Conservation analysis of rhamnose and RPS genes").
Pipeline as described in Methods (verbatim params):
- Query = 19 genes of strain P1/7 (NCBI NC_012925.1): 5 dTDP-rhamnose
(rmlA=SSU1133, rmlC=SSU1132, SSU1131[hypothetical], rmlB=SSU1130, rmlD=SSU1129)
- 14 RPS cluster genes (srpB=SSU1124, srpC=SSU1123, srpD=SSU1122, srpE=SSU1121, srpG=SSU1120, srpI=SSU1119, srpJ=SSU1118, srpK=SSU1117, srpL=SSU1116, srpM=SSU1115, srpN=SSU1114, srpP=SSU1113, srpQ=SSU1112, srpR=SSU1111). Locus-tag map from Table S1. (srpO=SSU1672 is a separate gacO ortholog, not in the 14-gene cluster rings.)
- Genome set: 1800 public S. suis genomes collected from public databases per the method of ref 28 (Roodsant & Van der Putten 2021, Virulence; PMC8632099), QC-filtered with QUAST v5.0.2 (drop >500 contigs / N50 < 10 kbp / size outside 1.6–3.0 Mbp) → 1719 retained.
- Screen: ABRicate v1.0.1, initial 60% cov / 80% id, then stringent 90% cov / 80% id. Conservation index = coverage × identity. Grouped per genome with ST, CPS serotype, host, geography; visualised as iTOL heatmaps with the tree.
Reproducible claims (see claims.tsv for verbatim text + location)
- C1 1800 collected → 1719 retained after QC.
- C2 dTDP-rhamnose genes ubiquitously present in all genomes, little variation.
- C3 ST-1 & ST-7: all RPS genes >91.5% conservation vs P1/7.
- C4 ST-16 & ST-20: <86.5% conservation in srpC & srpL vs P1/7.
- C5 RPS-cluster extraction (ABRicate 90/80 → extract_genes_ABRicate) → 1350 RPS sequences identified (then 95/80 stringent for alignment).
Also in scope but heavier (Prokka v1.14 → Roary v3.12 → MAFFT v7.307 → SNP-sites v2.5.1 → IQ-TREE v1.6.6 GTR+F+I+G4, 1000 bootstraps): the core-genome ML tree used as the iTOL backbone. Optional 20% — tree topology is not given as a numeric claim, so it is low-value to grade; deferred.
OUT OF SCOPE (wet-lab / manual / not pipeline-derived)
Lectin binding (Fig 2A/C), GC-MS glycosyl composition & linkage (Fig 2B/D, Tables S2/S3), NMR/GroP, mutant growth/morphology/TEM (Figs 4–5), lysozyme/complement/ hGIIA assays, antibody binding / opsonophagocytosis / vaccination (Figs 3,6).
KEY REPRODUCIBILITY LIMITATION (provisional)
The exact 1719-genome accession list is NOT deposited — absent from the supplement (0 GCA/GCF accessions in sciadv.adz1854_sm.pdf), from Zenodo (helper script only), and the data-availability statement is generic boilerplate. So the identical genome set cannot be reconstructed. We therefore reproduce by re-collecting all public S. suis assemblies from NCBI, applying the same QUAST QC, and testing the per-lineage conservation claims (C2–C4) — which are lineage properties robust to the exact set (ST-1/7/16/20 are common). C1 and C5 are snapshot-dependent and will be graded as partial/approximate. The query DB (C-core) is exactly reconstructable from P1/7 NC_012925.1 + Table S1.
Assessments & scoring basis
Each contributor’s verdict, the per-question basis, and the auditable, itemised worksheet behind it.
An automated assessment. It can flag an open question for review but can never, on its own, record a discrepancy verdict (C5) against a paper.
Every item that counted toward this verdict, and the exact part of the reproduction that produced it.
This is a P16 third-party-tool (ABRicate) reproduction of the single in-scope bioinformatic result; all wet-lab claims are out of scope. The central scientific findings reproduce cleanly from an independently collected public genome set — rhamnose genes ubiquitous/conserved (C2), ST-1/ST-7 RPS >91.5% (C3), and srpC/srpL as the divergent pair in ST-16/ST-20 with srpL clearly <86.5 (C4), srpC only 0.06pp over threshold. The only failures are the exact counts C1=1719 and C5=1350, which are not byte-reproducible because the authors' precise genome accession list is undeposited and our self-chosen 2026 GenBank snapshot is ~5x larger. This is a reproducibility/transparency gap on the data-availability side plus a self-defined-sample effect on ours — not fabrication; the core conclusion holds, so overall this is a solid yellow.
Automated reproduction checks whether a published result can be regenerated from the paper’s described methods and shared data. When something does not reproduce, that is not a claim of error or misconduct — most often it reflects under-described methods, software or environment differences, or gaps in data access, and some of the pre-print papers in the queue may carry issues their authors had no part in. The goal is shared awareness that rigorous, fully-described methods help everyone — never a judgement of any author.
Are you an author? We would genuinely like to hear from you — to clarify the record, add data or code, re-run the pipeline after an accession update, and publish your response right next to the assessment. Everything here is open and auditable.
🚩 Report an error in this record
Spotted something wrong — a verdict you’d contest, a data or value error, or a private detail that slipped through? Tell us, with a short justification. Authors and readers are equally welcome to write in; we review every report.
Prefer email, or the form below not working? Contact us at support@doesitreproduce.com.
Reproduction footprint
claude-opus-4-8Measured resources invested to assess this paper — sanitised (machine class only, no job ids/paths). Compute = HPC accounting (SLURM); tokens = the AI agent's session.