EEB Daily

Vol. 1, No. 33Wednesday, November 18, 2026Bryan, Texas
One paper worth the coffee. One paper from beyond the fence. One story from the wider world.

The lead · Phylogenetic methods for cell-lineage timing with state-dependent mutation rates

Cell family trees can tell time without one universal clock

A new Bayesian method dates immune-cell ancestry while allowing mutation rates to change with cell state.

A family tree made from DNA normally measures branches in accumulated mutations. Converting those branches into elapsed time requires a molecular clock, an assumption about how quickly mutations accumulate. That assumption can be troublesome inside one body. Immune B cells acquire mutations during their response to pathogens, and mutation rates can depend on the cells' identities or genetic backgrounds. Jessie Fielding and colleagues introduce TyCHE, short for Type-linked Clocks for Heterogeneous Evolution, to estimate dated ancestry trees while letting evolutionary rates depend on discrete states such as cell type. The method uses Bayesian phylogenetics, which combines a model of evolution with uncertainty about tree histories.

The team tested the approach with SimBLE, a new simulator of B-cell lineage evolution, and reports improved accuracy in estimated time trees. They then applied it to memory B-cell differentiation during HIV infection, germinal-center responses after influenza vaccination, a glioma tumor lineage and bacterial infection. These applications illustrate a broad methodological point: different branches of one evolving population need not share a single mutation rate. TyCHE and SimBLE are available as open-source software compatible with BEAST2 and the Immcantation analysis ecosystem. The accepted manuscript reports the broad findings, not a universal guarantee of accuracy across datasets.

Where to pushA direct next test would simulate cell states that change faster than they can be sampled, then ask whether TyCHE still recovers correct dates. Compare its estimates with independently timed lineage experiments rather than only reconstructed trees. Can state-dependent clocks distinguish true migration between tissues from an unrecognized shift in mutation rate?

Beyond EEB · A paper from outside evolution and ecology

Brain immune cells help erase practiced movements

Hu et al., Nature Neuroscience (17 Nov 2026). Research article.

Learning a movement involves changes to myelin, the insulating material wrapped around nerve fibers. In mice, Hu and colleagues find that declining motor memory is accompanied by microglia, the brain's immune cells, engulfing myelin sheaths associated with that memory. Reduced activity in memory-linked neurons lowers CD47, a molecular 'do not eat me' signal on myelin-producing cells. Removing CD47 accelerated both myelin loss and motor-memory decay; increasing CD47 had the opposite effect. The experiments suggest that forgetting can involve active cellular removal, not just fading neural signals. They do not establish that the same mechanism explains human forgetting.

Beyond biology · A result from outside biology

Melting in two dimensions is a contest between energy and disorder

Li, Ding and Ma, Communications Physics (16 Nov 2026). Research article.

A flat crystal need not melt in a single jump. Two-dimensional systems can pass through a hexatic phase, where particles lose long-range positional order but retain some directional organization. Li and colleagues analyze why melting sometimes happens gradually and sometimes through a sudden transition. Their framework separates energy, which favors some arrangements, from entropy, which counts possible arrangements, by examining the curvature of free energy. They identify conditions under which entropy favors an abrupt transition, and distinguish vibrational from configurational contributions. This is a proposed thermodynamic explanation of phase behavior, not proof that every real nanoscale material melts identically.