Haha, that's an interesting reframing @jonny ! Any useful concepts from cryptography we can import into neurobiology?
@manisha @ansuz
Re: Cell-type connections: clustered Protocadherin (cPcdh)-mediated self- and other-avoidance has been reported for dendrites in the retina and cerebellum, both regions where this tiling structure is required for the circuitry to function properly.
cPcdh randomization is controlled by epigentic modulation (histone methylation) ... (1/n)
... and one of the first things to go wrong when the 3D chromatin structure is disturbed, see e.g. here https://actaneurocomms.biomedcentral.com/articles/10.1186/s40478-022-01478-6
and here: https://academic.oup.com/nar/article/46/10/4950/4937548
It gets wilder in the cortex: There, cPcdhs regulate axon -> dendrite connection formation in a way that tracks how closely related the cells are.
The cortical plate has several generations of progenitors, each inheriting and partially remixing the cPcdh pattern. (2/n)
The final progenitor generation produces a set of a couple "sister cells" which preferentially form synapses with each other https://bmcbiol.biomedcentral.com/articles/10.1186/s12915-016-0326-6
In contrast, the less-related "cousin cells" set (likely to be direct surrounding neurons of the "sister cells") is contacted *less* than would be expected. This maintains a diversity of input processing, which is disrupted when cPcdh expression is artificially constrained: https://doi.org/10.1016/j.celrep.2023.112170
(3/n)
cPcdhs also seem to control where cortical neurons establish themselves in the cortex, i.e. how closely or far apart from their "sister cells" they settle before they make connections: https://www.nature.com/articles/s41586-022-05495-2#Abs1
So even though cPcdhs are not the only cell-surface molecules that control tiling/connectivity (that's a plausible hypothesis for each large family of signalling molecules), they by now have a relatively large body of evidence.
(4/n)
I also learned a lot, thanks for writing up the posts!