Thread by @drmichaellevin

Excited to share this preprint, with student @niwhskal : arxiv.org/abs/2207.14729. Genotypes mostly do not code directly for phenotypes, ok. But what implications for evolution of the *competency* of cells, within the developmental physiology that forms the “hidden layers” between

genome+environment (input) and form+function (output)? We analyzed minimal simulations of evolution of virtual embryos where cells have different levels of local autonomy prior to fitness evaluation (like reviewed in mdpi.com/1099-4300/24/6…). Lots of interesting aspects, but key

is: the more cellular competency, the less well evolution can see the genome, which reduces ability to select good genotypes – vicious cycle. What happens then? Most of the effort goes into increasing the competency. A sort of evolutionary ratchet for #BasalIntelligence. Also, an

explanation for planaria – the animal with the best anatomical control but a really chaotic genome, and unlike every other model species, no available mutant strains with altered morphologies (except our 2-head line (sciencedirect.com/science/articl…), which is of course not genetically

induced). I think these simulations show what may have happened there – the intelligence ratchet went really strongly, and evolution put most of the work into optimizing the algorithm that enables reliable morphogenesis despite highly variable genetic inputs. More on this coming.