Receptor complements as the unit of modulatory computation: two open problems

Most signalling molecules that act on cortex by diffusion bind several receptor subtypes, and those subtypes couple to distinct and often opposing intracellular pathways. The effect of such a molecule on a neuron is therefore set by the neuron's receptor complement: which subtypes it expresses, in what combination, and in which subcellular compartment. Comparative studies show that mouse and macaque achieve functionally equivalent cholinergic modulation of primary visual cortex through structurally different arrangements of receptor subtype and cell class. We hypothesise that the complement, rather than the molecule or the cell type, is the appropriate unit of description. This note records two open problems that follow, and the approach proposed for each, in advance of analysis. The first is localisation: predicting, for every cortically expressed receptor for a diffusible signal, the subcellular compartment to which it is trafficked. The second is combination: determining whether the pattern of receptor co-expression across neurons carries structure that the transcriptomic cell-type taxonomy does not already capture, and whether that structure is estimable from atlas-scale data under realistic detection. Both problems can be addressed from public data.