A 20 micrometre control reduced HVC route excess from 98% to 1.2% in one songbird Area X volume

We tested whether an HVC axon fragment and an MSN cell tend to contact the same GPi cell. Such wiring could support opposing direct and indirect effects. We used one adult male zebra finch Area X volume, j0251, with v8 cell labels. The main test used 43 GPi cells, 8,096 eligible MSN cells and 23,094 selected HVC-to-MSN synapses. Each input synapse and each direct GPi target of its source defined one route opportunity. We found 6,215 closed opportunities out of 25,391, or 24.48%. The exact shuffle mean was 12.37% across the whole volume and 24.18% within 20 micrometre cubes. The observed-to-mean ratio thus fell from 1.979 to 1.012. The local enrichment test gave p=0.094 from 999 shuffles. The HVC ratios were 1.006 at 10 micrometres and 0.999 at 5 micrometres. A 40 micrometre control left a 3.8% excess. A shifted 20 micrometre grid gave a 1.7% excess. A separate LMAN test gave a small 1.1% excess at 20 micrometres (p=0.001). Counts of distinct route triples were below the local shuffle means for both inputs. Local input counts account for most of the large excess seen with a global control. Small residual effects depend on spatial scale and the count rule. The test does not establish synaptic timing or measured inhibition. One brain, automated cell labels and a null that does not keep parallel connection counts limit the result. The package gives code, source hashes, exact null means, all test results, 63,891 derived candidate routes, figures and repeat instructions.

songbird · connectome · Area X · feedforward inhibition · spatial control · synaptic motifs · HVC · LMAN