What'sNEW

11 Mar 2025

What'sNEW about HGT

For review, HGT is "all there is" for the evolution of prokaryotes. (Ernst Mayr said these words to me 25 years ago.) Extensive, undisputed evidence now supports this statement, but its significance has not sunk in. The evidential tipping point, for me, was a computer analysis that attempted to match the graph of gene distribution in E. coli. The graph showed that a few genes were always present, the "core genome," but far more genes were only found once, the "pan-genome." The successful computer model ignored duplication-and-divergence and de novo gene creation, and relied on HGT exclusively.

Now Chinese biologists have analysed the genomes of different strains of a worm that threatens agriculture. They divide the worm genes into four categories instead of two, and the core group is a much larger fraction of the total than was found in E. coli. But the story is similar. There's even a U-shaped graph. I suggest that, for eukaruotes as well, in order to acquire new genes, HGT may be "all there is."

They note that the subunits have features that could not have been initially acquired because of selection for the functions of the final complex, because they existed before those functions first appeared. In spite of this hurdle, they they close with reassurance, Although evolution of any particular molecular sequence or architecture without consistent selection for those properties is vanishingly improbable, our findings suggest that proteins evolve constantly through a dense space of possibilities in which complex new interactions and functional states are easily accessible.

What are they saying here? The gene for each of the subunits of hemoglobin is over a thousand nucleotides in length. Composing a gene that long by point blind mutations is vanishingly improbable, as Pillai et al. acknowledge. What they seem to be saying is that if the complex subunits are available, they can be properly pieced together by mutation and selection in only a few steps.

Cosmic ancestry says the same thing. If complex programming is installed in a few, large, compatible pieces, these can be pieced together and made functional by a genetic operating system. (I call it "robust software management.") Very many likely examples are known — that's why horizontal gene transfer (HGT) is effective. And the process would be easy to demonstrate in computer models. For example, a thousand-word text arbitrarily fragmented into half-a-dozen pieces could be reassembled with the syntax- and spell-check logic in a word processor. But it's a very different theory of evolution.