One-site signals retained 51 to 88 percent of mean linear information in eight partial worm circuit models

We tested whether high linear information retention for a common synapse-site signal also holds for a signal at one site. We used the same 162-neuron panel in eight public C. elegans maps. Independent sites had a common baseline release probability and unit target effects. For each source, the exact mean retention over single-site signals is the rank of the site-to-target matrix divided by its site count. Equal-source map means were 50.58 to 87.94 percent, versus 98.14 to 99.71 percent for common signals. The two adult maps retained 50.58 and 51.61 percent for one-site signals. Site-weighted means were 46.09 and 44.93 percent in those maps. Exact state checks confirmed the linear calculation and showed that zero linear information need not mean zero full distribution information. These are model results, not measured release or coding. Site-specific signals, independent release, unit effects, ideal target access and partial brain coverage limit inference. Eight animals in one species cannot establish an age effect or cross-species transfer.

connectomics · linear information · site-specific signal · C. elegans · polyadic synapses · model limits