5,800 genomes, one map: Vienna team says chromosome change runs one way only

The load-bearing claim is about an operation with no inverse. When two chromosomes fuse and their genes intermingle, the mixing cannot be undone, and the event stays in the genome as a permanent record ^[4]. That asymmetry is what lets genome shape do evidential work rather than illustrative work: because the changes run in one direction only, they mark shared ancestry, and the authors note such evidence has already been used to identify the sibling group to all other animals ^[5]. The same logic explains why a chromosome count is a poor summary on its own. Differences in chromosome number between animal groups come either from ancestral chromosomes combining or from their separation, and in both cases the mixing sends lineages down very different routes ^[6].

The numbers deserve a second look. More than 5,800 assemblies cover 4,454 species ^[2], which works out to about 1.3 assemblies per species, meaning roughly 1,350 genomes, close to a quarter of the set, add depth on species already represented rather than new ones ^[12]. Spread across 19 phyla, the species count averages 234 per phylum ^[13]. That is a real map and also a coarse one, since a coordinate is only as trustworthy as its phylum's sampling, and the highways themselves are described as being traced through hundreds of present-day species that got on or off them at different times and rates ^[11].

The admission requirement is strict. Draft genomes list which genes an animal has but not their order, while chromosome-scale assemblies place every gene along complete chromosomes, are much harder to produce, and have only recently reached numbers that permit a comparison across the animal kingdom ^[8]. Adding a thinly covered phylum to this map is therefore a sequencing decision, not an analysis decision.

Prediction is the part to hold lightly. Simakov's claim is that understanding these rules lets researchers ask where genome evolution might go next and identify measures for conserving animal biodiversity ^[14]. The study's stated uses are to prioritise unusual lineages for deeper study and to test whether chromosome changes are linked to shifts in gene regulation, development or biodiversity ^[15]. Those are tests the framework makes available, not results it reports. What is reported is that progressive one-way mixing places major groups in distinct regions of genome architecture space and leaves an imprint on a broad range of genes, developmental control genes among them ^[7], and that thousands of genomes can be seen on one map and refolded to compare groups after they diverged, in the account of Darrin Schultz, who led the work at Vienna and is now at Lehigh ^[9]. The deep-time anchor is why any of this holds at all: a human and an octopus still carry recognisable pieces of a genome inherited from an ancestor that lived more than 600 million years ago ^[10].