01 Science · 24 September 2026_
Ancient introgression drives wild boar expansion and phenotypic diversification of domestic pigs
- Population genetics
- Animal science
745 Eurasian Sus genomes place the wild boar expansion in the Middle Pleistocene from South Asia, preceded by a Pliocene introgression that brought in a 43 Mb X-chromosome haplotype later selected for body size in domestic pigs.
Summary
The authors analysed 745 genomes from wild boar and domestic pigs across Eurasia. Wild boar expanded out of South Asia during the Middle Pleistocene. Before that, about 3 to 5 million years ago, an archaic Suinae lineage introgressed into the ancestral wild boar population in South Asia and contributed a haplotype of about 43 Mb on the X chromosome. The haplotype contains loci associated with larger body size and cold adaptation. During domestication it was selected at least three times independently in the development of large-bodied breeds.
What to take from it
- 745 genomes, Eurasia-wide sampling
- Wild boar expansion from South Asia, Middle Pleistocene
- Pliocene introgression from an archaic Suinae lineage, ~43 Mb X haplotype
- Haplotype selected at least three times independently in large-bodied breeds
Review
A 43 Mb introgressed block surviving millions of years on the X is a strong claim; low recombination on the X, or an inversion, would explain its persistence and should be tested directly with long-read assemblies of carriers and non-carriers. Dating an introgression at 3 to 5 Mya depends on the mutation rate and generation time used and on which archaic lineage is taken as donor, which the abstract does not name. "Associated with body size" needs the association design: X-linked loci are hard to map in pigs because of sex-specific dosage and breed structure. Before using the haplotype as a breeding marker: its structure in assemblies, whether candidate genes within it have functional support, and whether the three selection events are independent or one sweep spread by crossbreeding.
02 Genetics Selection Evolution · 21 September 2026_
The rate and spectrum of germline mutations in chicken from a commercial pedigree line
- Animal science
- Population genetics
Whole-genome sequencing of 194 chicken trios gives a germline mutation rate of 3.49 x 10^-9 per site per generation, a 2:1 male bias, and a gene ranking by expected contribution to mutational variance.
Summary
The authors sequenced 194 parent-offspring trios from one commercial broiler pedigree line. The estimated rate was 3.49 x 10^-9 mutations per site per generation, about 6.2 new mutations per offspring, and counts fit a Poisson distribution. Read-based phasing assigned about twice as many mutations to fathers as to mothers. Microchromosomes had a point estimate about 8% higher than macrochromosomes, consistent with higher GC content and a spectrum dominated by C to T transitions. From the spectrum they built a codon transition matrix and ranked genes by their chance of non-silent mutation; immunity and development genes ranked high.
What to take from it
- 194 trios, one commercial line
- mu = 3.49 x 10^-9 per site per generation; ~6.2 de novo mutations per bird
- Paternal:maternal ratio about 2:1
- Gene-level ranking of expected contribution to mutational variance (Vm)
Review
A trio design this size gives a solid direct estimate for one line. About six mutations per offspring is few, so the callable-genome fraction and false-negative rate set the estimate; those need simulation or validation numbers, and microchromosomes are where short-read callability is worst, so the 8% difference is within what callability alone could produce. The Vm ranking assumes mutational effect follows from codon composition and target size, not from observed effects on traits; it is a prior, not a measurement. Before using the rate in demographic dating of chicken history: parental ages, since generation interval in a commercial line differs from wild or village chickens, and replication in a second line.
03 Journal of Experimental Botany · 26 September 2026_
Haplotype-specific analyses in apple: insights from 'Golden Delicious'
- Methods
A phased telomere-to-telomere 'Golden Delicious' assembly, placed in a six-haplotype gene pangenome, lets the authors measure allele-specific expression, with one haplotype dominating expression of the fruit-softening gene Md-PG1.
Summary
The two haplotypes of 'Golden Delicious' assemble to 647.3 and 649.2 Mb. Phasing was checked with 10,321 SNPs and chromosome ends were confirmed by telomeric repeats. Gene prediction gave 45,116 and 45,063 genes. A gene-family pangenome of six haplotype-resolved apple genomes separated shared and unique families. The phased reference allowed genome-wide allele-specific expression; for Md-PG1, the haplotype 2 allele contributed most of total expression. The haplotypes also differed in miRNA distribution and methylation.
What to take from it
- Phased T2T assembly of a founder cultivar, ~648 Mb per haplotype
- Six-haplotype gene-family pangenome
- Genome-wide allele-specific expression; Md-PG1 driven by haplotype 2
- Haplotype-level miRNA and methylation differences
Review
The value is the resource: a phased founder genome makes allele-specific analyses possible for breeding. The pangenome is gene-family based on six haplotypes, so it describes little of apple diversity and says nothing about structural variants between haplotypes. Allele-specific expression from RNA-seq mapped to a diploid reference is biased toward the better-assembled or more similar haplotype unless mapping bias is controlled; the Md-PG1 result needs that control and a second tissue or time point. Validation of phasing with 10,321 SNPs is thin for 650 Mb. Before building on it: switch-error rate from trio or Hi-C data, and a graph of SVs between the six haplotypes.
04 Journal of Computational Biology · 24 September 2026_
Interpolating and extrapolating node counts in colored compacted de Bruijn graphs for pangenome diversity
- Methods
Node counts in colored compacted de Bruijn graphs are made comparable across pangenomes built from different numbers of genomes, using rarefaction-style interpolation and extrapolation plus Hill numbers.
Summary
Pangenome graphs built from different numbers of genomes have node counts that cannot be compared directly, and rare sequences inflate the counts. The authors treat each node in a colored cdBG as a unit and the set of genomes traversing it as its frequency, then borrow two tools from ecology: interpolation and extrapolation of counts to a common number of genomes, and Hill numbers, which weight rare and common nodes by their frequency.
What to take from it
- Target: colored compacted de Bruijn graphs
- Rarefaction and extrapolation of node counts to a common sample size
- Hill numbers to control the weight of rare nodes
- Transfers species-diversity statistics to pangenome graphs
Review
A useful reframing: pangenome openness as a species-accumulation problem, with a statistical toolkit already tested in ecology. Node counts in a cdBG depend on k and on sequencing errors, so comparisons across graphs need the same k and error filtering, and results should be reported across k. Hill-number extrapolation assumes genomes are sampled independently from one population; clonal or structured sampling, the usual case in bacteria and livestock, biases it. The abstract gives no benchmark. Before using it: behaviour on simulated pangenomes with known openness, sensitivity to k, and comparison with Heaps' law estimates on the same data.
05 PLoS Pathogens · 25 September 2026_
A bacterial pangenome association study identifies a role for polyamine metabolism and alkaline tolerance in middle ear infections
- Microbial genomics
A phylogeny-aware pangenome association study of over 200 Haemophilus influenzae isolates finds that losing the potE-speF polyamine operon is linked to otitis media, and knockouts raise middle ear bacterial burden about 63-fold in mice.
Summary
The authors sequenced over 200 H. influenzae isolates collected over time from 9 healthy and 12 otitis-prone children, clustered genes into a pangenome and identified clonal lineages. Phylogeny-informed tests found 7 strong gene or operon associations with health or disease; 5 were gene absences associated with disease, all involved in pH homeostasis and polyamine transport or modification. In vitro, potE-speF made cells sensitive to alkaline pH in the presence of ornithine. Middle ear fluid was more alkaline when H. influenzae was present. In a mouse model, deleting potE-speF raised middle ear burden about 63-fold, with no difference in a lung model.
What to take from it
- >200 isolates from 21 children, longitudinal
- Phylogeny-aware gene presence/absence association
- 5 of 7 hits are gene losses linked to disease, all pH or polyamine related
- potE-speF deletion: ~63-fold higher middle ear burden in mice, no lung effect
Review
The strength is the chain from association to knockout to animal model, which few pangenome association studies complete. The sample is 21 children with repeated isolates from the same lineages, so the effective sample size is much smaller than 200 and associations can reflect a few hosts; how the tests handle host as a unit matters more than the phylogeny correction. Gene absence calls from short-read assemblies can be false when genes sit at contig breaks. The pH measurement in human fluid is correlational. Before generalising: replication in an independent cohort, and whether potE-speF loss is enriched in otitis isolates in public H. influenzae collections.