EEB Daily

Vol. 1, No. 29Saturday, November 14, 2026Bryan, Texas
One paper worth the coffee. One paper from beyond the fence. One story from the wider world.

The lead · Meiotic recombination and GC-biased genome evolution in ants

Ant chromosomes shuffle at extreme speed

A new map of California harvester ant chromosomes links exceptionally frequent crossovers to genome diversity and GC-rich DNA, without clear evidence of stronger protein-level selection.

Ants, bees and wasps can reshuffle their chromosomes at extraordinary rates during meiosis. Mohammed Errbii and colleagues built a linkage map for the California harvester ant, Pogonomyrmex californicus, using DNA markers scored in 105 haploid males produced by one queen. Because each male carries a single maternal genome, switches between marker states reveal crossovers, the exchanges that occur between homologous chromosomes. The team anchored roughly 95% of the genome to 16 chromosomes and estimated a genome-wide recombination rate of 14.73 centimorgans per megabase, the highest reported for an ant.

Recombination was scarce in repeat-rich, gene-poor stretches but increased in gene-rich regions. More recombining regions also showed greater nucleotide diversity and higher GC content. At coding sites, rates of silent substitutions fell as recombination rose, while GC at third codon positions increased, a pattern consistent with GC-biased gene conversion favoring G and C bases during DNA repair. Yet local recombination was not robustly associated with protein-changing evolution or an estimate of selection efficacy. The findings therefore separate strong compositional effects from harder-to-detect effects on protein evolution. Crucially, these are correlations within one species, not experiments proving that recombination causes the patterns.

Where to pushDo high crossover rates change how efficiently selection removes harmful variants, or mainly reshape base composition? Compare multiple ant species with independently measured maps, similar demographic reconstructions and matched estimates of coding and noncoding diversity. Within species, tests that account for variation in mutation rate and linked selection could distinguish GC-biased repair from adaptive substitution.

Classic · A landmark paper, revisited

Recombination frees genes from each other's fate

Felsenstein, Genetics (1 Oct 1974). Original paper.

Why should sexual reproduction survive when it repeatedly breaks apart successful combinations of genes? In 1974, Joseph Felsenstein revisited a dispute between models that did and did not predict an advantage for recombination. His key distinction was population size: in finite populations, random genetic drift creates associations between linked alleles, even when the genes do not directly interact. Selection at one site can then interfere with selection nearby. Through analytical arguments and computer simulations, Felsenstein showed how recombination can relieve this interference and slow the buildup of harmful mutations, connecting the Fisher-Muller argument to Muller's ratchet. Later theory expanded how we study interference among linked sites.

Beyond biology · A result from outside biology

A circuit compiler cuts the cost of quantum error correction

Kuehnke et al., npj Quantum Information (12 Nov 2026). Research article.

Quantum computers encode information redundantly to catch errors, but error correction also makes even routine logic operations expensive. Kuehnke and colleagues developed a mathematical compiler that designs physical circuits for logical Clifford gates, an important class of quantum operations, while respecting which qubits a device can connect. Their optimization searches equivalent implementations and minimizes two-qubit gates. In tests on small error-detecting codes, the new circuits used fewer operations than a standard Qiskit approach. Simulations of another small code predicted improved logical-error performance, sometimes by more than fourfold compared with a teleportation-based design. These are software and simulation results, not a demonstrated improvement on a working quantum processor.