BGPT: Paper Review: House Mice in the Atlantic Region: Genetic Signals of Their Human Transport

Explore by Goal

Quick Explanation Copied

What the paper claims (testably): using mitochondrial D-loop haplotype phylogeny (2,765 mice; 758 haplotypes; 47 countries/oceanic archipelagos) the authors infer broad, human-mediated Atlantic colonization routes for Mus musculus domesticus, with Northern Europe as a major source for many Atlantic regions and evidence for additional contributions from Southern Europe, Scandinavia, and Macaronesia (including a “Viking transport” interpretation).

Long Explanation

Paper Review (Critical, Visual, Evidence-based)

House Mice in the Atlantic Region: Genetic Signals of Their Human Transport

DOI: 10.3390/genes15121645

1) What they did (methods mapped to inference)

Marker: mitochondrial D-loop haplotypes, maternally inherited; the authors explicitly use it as a “broad-brush” phylogeographic signal.
Data assembly: new sequences plus published data, totaling 2,765 mice and 758 haplotypes across 47 countries/oceanic archipelagos.
Phylogeny: rooted maximum-likelihood phylogeny from haplotypes; the paper reports previously named clades B–F and uses their prevalence patterns for source inference.
Inference logic: comparisons of clade composition and shared multi-location haplotypes between source and colonized regions using chi-squared association tests and haplotype-affinity proportions.

Skeptical framing: all inference maps to “human transport” only if alternative explanations (natural dispersal, post-colonization reshuffling, demographic replacement) cannot plausibly generate the same mtDNA pattern. The paper acknowledges limitations and treats the results as broad-scale signals rather than precise temporal trajectories.

2) Visualizing their key contingency tables (from the paper)

The paper reports clade counts (Table 1) and chi-squared p-values (Table 2). Below, I plot the clade-by-region haplotype and clade-by-region individual composition that underpins their source-matching logic. Values are transcribed directly from the provided Table 1 in the text you supplied.

Critical note: Inference uses both haplotype counts and individual counts, but mtDNA haplotype frequency is also shaped by post-colonization demography and possible sampling imbalance (e.g., some regions have sparse sampling). The paper flags that some areas (e.g., Caribbean) have insufficient data for inference in their Discussion.

3) Chi-squared source matching (Table 2) — visual p-value magnitude

The paper reports chi-squared p-values comparing source regions vs colonized regions. Below is a heatmap of the p-value magnitudes transcribed from Table 2(a) and 2(b) blocks shown in your supplied text (the authors report extremely small p-values in many comparisons).

Interpretation constraints: p-values depend on sample sizes and the contingency table degrees of freedom; the authors also merged clades to satisfy test assumptions, which can mask differences.

4) Claimed historical signals (what is plausible vs what is a leap)

A. Northern Europe ↦ Scandinavia / North Atlantic islands

The authors report that North Atlantic islands have the closest clade characteristics to Scandinavia, and that Scandinavia as a source for North Atlantic islands matches chi-squared results and Viking-transport expectations.

B. Macaronesia as a stopover with multi-source impacts (incl. Viking transport signal)

The paper claims Macaronesia displays similarity to Southern Europe but also a signal interpreted as Vikings transporting mice from Germany/Scandinavia.
They further claim Macaronesia impacted South Atlantic island colonization and Latin America’s mouse genetic makeup, via dashed vs dotted arrows in their inferred route figure.

Skeptical critique: multi-source mtDNA patterns can arise from repeated introductions, but they can also emerge from later mixing after establishment. With mtDNA alone, distinguishing founding introductions from post-founding replacement is inherently difficult; the paper frames results as broad-scale signals rather than precise timing.

C. Specific haplotype-sharing examples (azores↦falklands; dom2 as a British-marker candidate; Tristan as Portuguese)

Example: dom650-652 shared between Azores and Falkland (Malvinas) within a larger group interpreted as supporting Azores as a colonizing source for Falklands.
They present dom2 (clade E) as a candidate marker of British colonial history, but explicitly caution that dom2 could also have been introduced by other countries/visitors (France/USA) and that the inference is probabilistic.
Tristan da Cunha: they report first D-loop sequences for the island system and argue mitochondrial signatures likely reflect Portuguese visitation rather than later British colonization, using the relationship of Tristan haplotypes to Portugal and Brazil in clade structure.

5) Blind spots & counterpoints (what could falsify the inferred routes)

Marker limitation (mtDNA-only): mtDNA tracks maternal lineages; colonization can involve male-biased transport or later demographic replacement, which could decouple mtDNA signals from colonization routes. The paper mitigates by emphasizing broad-scale patterns and multi-location haplotypes, but mtDNA still cannot resolve genome-wide admixture complexities.
Sampling imbalance & geographic gaps: some regions are under-sampled to the point that colonization history cannot be inferred (e.g., Caribbean). This can bias “closest” sources simply because some haplotypes were not sampled.
Chi-squared contingency design: clade merging to satisfy test assumptions can mask differences; the selection rule favors higher chi-squared, which may systematically bias which source is “best.”
Correlation vs causation for “human transport”: the paper interprets haplotype/clade similarities as consistent with maritime routes, but genetic similarity can also be produced by alternative historical pathways (e.g., undocumented stops, ecological corridors, or stochastic drift in bottleneck events). The authors’ strongest claims are those that align multiple independent genetic signals (clade composition + specific multi-location haplotypes) and acknowledge uncertainty in others (e.g., dom2 alternative introductions).

6) Data availability & reproducibility posture

New sequences deposition: the authors state all new DNA sequences are added to GenBank with accession numbers PP751342–PP751400.
Supplementary materials: the paper provides a link to supplementary tables/figures including haplotype catalogs and chi-squared values.
Reproducibility limitation: the paper provides methods and tool versions but your supplied text does not include the full alignment/haplotype-by-sequence matrix; that is typically necessary for fully reproducing the phylogeny. However, because accessions are provided and supplementary tables are described, partial reproducibility is likely strong.

7) Author review links (bespoke next step)

Feedback:

Updated: March 30, 2026