Where do the body’s axes come from? In Cognimed they are not imposed — they are the low eigenmodes of the embryo’s own electrical geometry.<\/p>\r\n\r\n\r\n\r\n
The operator.<\/strong> Cells are coupled by gap junctions into an electrical syncytium. That coupling defines a graph Laplacian — the embryo’s own diffusion operator — and its lowest eigenmodes are smooth standing patterns over the tissue: a head-to-tail mode, a back-to-belly mode, and a left–right mode whose node is the midline. These are the body axes. The anterior–posterior mode matches the real axis in the mouse atlas at r ≈ 0.96, and the frame is grown from the embryo’s own connectivity, never borrowed from the atlas.<\/p>\r\n\r\n\r\n\r\n The face and the organs.<\/strong> The same construction on the face yields the eigenframe on which facial features are placed — GWAS facial-shape variation is low-rank in exactly this basis — and each organ repeats the trick internally: the heart’s first eigenmode is its apex-to-base activation axis, the gut’s is the oral-to-aboral peristaltic axis. Organs and limbs are placed at the antinodes of the body-scale modes, addressed by the HOX code<\/a> on the anterior–posterior mode (derived from single-cell Hox expression, Spearman 0.81).<\/p>\r\n\r\n\r\n\r\n The amphibian threshold.<\/strong> The left–right eigenmode exists only if the body has medio-lateral width: a thin, convergent-extension-collapsed body has no such mode and grows no limbs — a fish. Widen the body (a single Wnt-PCP knob) and the mode enters the spectrum, and bilateral limbs appear at its antinodes — a tetrapod. The fish-to-tetrapod transition is a geometric threshold, with the limb genes unchanged.<\/p>\r\n\r\n\r\n\r\n