EXPERIMENTS
the experiments. every number on this page comes from a results file with a date on it. the rules were written before each run, and whatever came out is printed.
reading the results files
E2 · MEASURED
RESTING STATE
PASS RULE WRITTEN BEFORE THE RUN: ·
results file: ·
does the wiring alone predict which regions move together? the model runs with no stimulus, and the correlation of every pair of regions is compared with the real wiring and with a shuffled copy.
LIVEthe model at rest, developing · 0 s · r on half B, low-passed at 0.5 hz like the run, causal: ·(spearman vs SC; the run that passed gave · on ·)
drag the curtain across the plate: to the eye the two runs are nearly the same picture, and only the numbers separate them. move over either map to read a pair on both, click to hold it and light the two regions on the brain below.
| G chosen (on half A) | · |
|---|---|
| r fit, half A | · |
| r test, half B | · |
| r shuffled, half B | · |
| null p95 (1000 rewirings) | · |
| p | · |
| pairs · half A / half B | · |
| simulated · sampled | · |
move over the histogram to count the rewirings in a bar.
next: the same test against recorded chimpanzee brain activity, when that data is published.
E3 · MEASURED
LESION
PREDICTION WRITTEN BEFORE THE RUN: ·
results file: ·
switch one region off, with all its links, and measure what the rest of the brain loses: dGE is how much of the network's reach is gone, dFC how much the correlation of every remaining pair changed after a fresh run. all 244, one at a time. click a region to switch it off and read its row.
CLICK ANY REGION TO SWITCH IT OFFdrag to turn the brain
BRAIN INTACT
- GLOBAL EFFICIENCY
- ·
- dGE · MEASURED NOW
- ·
- dFC · FROM THE 60 S RUN
- ·
- WORST HIT REGION
- ·
the map prints dFC from the stored run. switch a region off and it reprints live: every other region inked by how much of its own reach it just lost.
TOP 10 BY dFC
| id | region | dFC | dGE |
|---|
·
·
·
E4 · MAPPED
DIVERGENCE
where the chimpanzee wiring differs most from the human one, painted from the atlas authors' own map (Figure 3, c2h minimum KL). this one is shown, not simulated: it is their measurement, on their surface, and it stays theirs.
source: c2h_minKL.L / R.func.gii, ChimpBNA SourceData/Figure3 (FANLabCASIA/ChimpBNA @ de5a36b), one value per surface vertex
10 MOST DIVERGENT REGIONS
| id | region | mean |
|---|
mean of the map over each cortical region's vertices where it is defined. the map covers the cortex.