In this intriguing article, the author explores how the genome encodes the form of an organism. Unlike simplistic metaphors like a blueprint or program, the author suggests a new analogy inspired by machine learning and neuroscience: that the genome encodes a generative model of the organism. This model consists of latent variables that specify the biochemical properties of proteins and regulatory factors, shaping an energy landscape that guides the self-organizing processes of development. By drawing parallels to Waddington’s epigenetic landscape, this analogy offers a fresh perspective on the genetic architecture of traits and the evolvability of developmental processes.
https://arxiv.org/abs/2407.15908