Check Claims Before You Publish

 Quick Explanation Copied

 Long Explanation

Evidence Base and Research Programme

Nadeau's recent work anchors a coherent, multi-layered programme on Heliconius adaptation. A 2025 preprint disentangles heritable adaptation from plasticity across altitude in H. erato and H. melpomene using a common garden (29–45 and 20–25 families, >800–1,549 individuals per trait class) plus reciprocal transplant at 19°C/24°C, finding high heritability of wing melanisation (h²≈0.99 in H. erato, ICC×2; p<1e-32) and parallel altitudinal clines consistent with thermal melanism Altitude adaptation vs plasticity preprint.

A 2026 sensory-ecology preprint shows heritable, elevation-associated divergence in facet count, optic lobe volume, and red filtering pigments (wild n=314 across 32 sites; F2 hybrids n=186), with PST exceeding genome-wide F_ST for eye traits and a facet-count × optic-lobe interaction on acuity (β=0.18, 95% CI 0.02–0.33) Heliconius eye and brain adaptation preprint. A companion 2026 developmental study uses CRISPR knockouts of ECM genes (tyn, min, Cpr128) to causally implicate extracellular-matrix dynamics in scale nanostructure formation ECM and scale nanostructure CRISPR preprint.

Citation Metrics and Track Record

OpenAlex records 90 works, ~5,279 citations, and h-index 29; her impact is front-loaded — the 2012 Nature genome paper (~1,214 citations) and 2013 speciation-with-gene-flow paper (~658) anchor her profile, and she is first author on the 271-citation cortex mimicry-gene Nature paper and a 317-citation genomic-islands paper cortex mimicry gene Nature paperGenomic islands targeted sequencing paper.

Critical Assessment

Strengths: methodological breadth (LMMs, PST-F_ST, Bayesian censored regression, CRISPR, TEM, omics), rigorous designs (common garden + reciprocal transplant; common garden + hybrids), and outstanding reproducibility — preprints deposit all data and scripts on Zenodo with named DOIs Zenodo data availability in altitude preprint. Limitations: the tyn bulge comparison used only 2 crispant vs 3 control individuals CRISPR sample-size limitations in ECM preprint; H. melpomene reciprocal-transplant power was small (>6 families); heritability may be inflated by maternal effects (authors note this); PST depends on assumed c/h²; and two key 2026 works are unpublished preprints with zero citations, so independent replication is pending. These are typical of honest, well-reported early-stage work rather than red flags.

What would change this assessment: peer-reviewed versions with independent replication of the CRISPR phenotypes in larger cohorts, and confirmation that altitudinal heritability estimates hold with maternal-effect controls (e.g., cross-fostering).

 BGPT Author Review

Scientific Quality

H-index 29 with ~5,300 citations across 90 works, including multiple Nature/Science first- and co-authorships, reflects an established evolutionary geneticist at the centre of the Heliconius genomics field. Recent preprints show methodological range (quantitative genetics, population genomics, CRISPR functional validation, sensory ecology) and transparent limitation reporting. Weaknesses: some knockout experiments rest on very small cohorts (2-3 individuals), and several high-profile claims await peer review and replication; no sign of fabrication or spin.

Communication Quality

Papers are well-structured, quantify uncertainty (CIs, mixed models), and report negative/ambiguous results (e.g., Cpr128 socket defects absent) — a hallmark of clear scientific communication. Deduction: some findings are split across dense supplementary analyses and preprints that are not yet peer-reviewed, and effect communication for non-specialists could be improved.

Author Novelty

Work tackles genuinely open problems — disentangling plasticity vs genetic adaptation in tropical altitude gradients, modular evolution of the visual system, and ECM gene roles in structural colour nanomorphogenesis. Not paradigm-creating at the level of her 2012 genome paper, but consistently generates first-in-field causal and quantitative-genetic evidence.

Scientific Rigor

Designs are exemplary: reciprocal transplant plus common garden to split G from E, PST-FST frameworks, preregistered-style Bayesian priors, null results reported, and complete data/scripts deposited on Zenodo with DOIs. Deductions: tiny crispant cohorts, uncontrolled maternal effects acknowledged but not resolved, and lab heating assays that may not capture natural solar radiation.