EEB Daily

Vol. 1, No. 9Saturday, October 24, 2026Bryan, Texas
One paper worth the coffee. One paper from beyond the fence. One story from the wider world.

The lead · Karyotype evolution: breakpoints and synteny decay

Mammal genomes break in pairs, and not just anywhere

Fitting three analytic models of chromosome breakage to synteny blocks from nearly 200 mammal genomes kills the random model outright and leaves two survivors that nobody can yet tell apart.

Comparative genomics runs on synteny blocks, stretches of chromosome in which gene order is preserved between two species. Whether a given block counts as surprising depends entirely on the null model: how are rearrangement breakpoints distributed along a chromosome? Since 1984 the default has been the random breakage model, with breakpoints falling uniformly, and for two decades the main challenge has been the fragile breakage model, which concentrates them in reusable hotspots. The dispute has mostly been fought over the shape of block-size distributions from a handful of genomes, and the answer sets the baseline for every claim that a conserved region, or a rearrangement, means something.

Moffett and Di Pierro derive exact block-size distributions for three models: uniform breakpoints, hotspots, and a new correlated breakage model in which breakpoints arrive in nearby pairs, which is what inversions and local rearrangements actually do. Fit to empirical distributions across nearly 200 mammalian genomes, the random model is decisively rejected. Hotspots and correlated pairs both fit well, and their inferred parameters recover rearrangement rates close to published estimates. Either one, the authors argue, is a better null than the model the field has been using.

Where to pushThe two survivors make the same prediction here, so block sizes alone cannot separate hotspot biology from rearrangement mechanics; breakpoint coordinates and pairing distances could. Worth noting too that nearly 200 mammal genomes are not 200 independent trials: pairwise comparisons share phylogeny, and the likelihoods treat that shared history as fresh data.

Classic · The paper that put trees in every regression

Felsenstein counts the data points, and there are fewer than you think

Felsenstein J. Phylogenies and the comparative method. The American Naturalist 125:1–15, 1985. doi:10.1086/284325

Correlating two traits across species had long been treated as ordinary regression with one observation per species. Felsenstein showed with a simple branching simulation that this is wrong: closely related species resemble each other because they share history, so the observations are not independent, and the resulting confidence intervals can be badly overconfident. His fix was independent contrasts: under Brownian motion on a known tree, differences between sister nodes give n minus 1 quantities that are independent with calculable variances. Everything downstream, phylogenetic generalized least squares, Ornstein-Uhlenbeck models, phylogenetic mixed models, grew from that argument.

Beyond biology · Number theory

A machine-written proof lands in number theory, and the checking has only started

Reported by Quanta Magazine, 19 Oct 2026; proofs posted by OpenAI 6 Oct 2026. quantamagazine.org

The Riemann hypothesis says every nontrivial zero of the zeta function sits exactly on the line with real part one half. The quasi version asks only that the zeros fall in a strip narrower than the classical bound of zero to one, and most consequences of the full hypothesis follow from it. In a batch of model-generated results, one paper proves a strip of one eighth to seven eighths across 199 pages and has been formally verified in Lean; a 49-page sequel tightens the strip to one twelfth and eleven twelfths and has not been. Specialists mostly believe the sharper result, call the writing poor, and are producing their own short versions; one says it is not a theorem until humans have checked every line.