FPL Survives 600 Million Years: Cross-Species Replication in the Drosophila Larva Connectome
Pre-registered cross-species FPL replication in the Drosophila larval connectome (n=2952 neurons, Winding et al. 2023): h1 CONFIRMED Pearson r=0.363 / Spearman rho=0.663, p=0.002. Hub rank order conserved across 600 million years of evolution. Spearman >> Pearson discrepancy is the primary finding. 25/31 pre-registered, p=0.000439.
The Strongest Test Yet
The Functional Proximity Law (FPL) states that hub importance scores in multilayer networks persist more strongly between functionally similar layers than dissimilar ones. Today we report F13: the FPL confirmed in the complete Drosophila melanogaster larval brain connectome, pre-registered before analysis (hash 7f04e5dd, commit fd5d315, 2026-05-26).
The Dataset
- Winding et al. (2023, Science 379, eadd9330) published the first complete connectome of a Drosophila larval brain. We used their Supplementary Data S1 connectivity matrices:
- axon_to_dendrite layer: 63,518 directed synaptic contacts (canonical chemical synapse direction)
- axon_to_axon layer: 40,173 directed contacts (non-canonical axo-axonic contacts)
- 2,952 neurons total, 103,691 edges combined
Layer taxonomy is from Winding et al., not from IRDME. Fully external dataset.
Results
| Hypothesis | Verdict | Key numbers | |---|---|---| | h1 FPL holds | CONFIRMED | Pearson r=0.363, Spearman rho=0.663, p=0.002, 95% CI [0.331, 0.394] | | h2 effect >= 0.40 | PARTIAL | Pearson r=0.363 < 0.40; Spearman rho=0.663 > 0.40 | | h3 divergents exist | CONFIRMED | 747 of 2,952 neurons divergent (25.3%) |
The Primary Finding: Spearman >> Pearson
The gap between Spearman rho=0.663 and Pearson r=0.363 is the primary scientific result. Hub rank order is strongly conserved across synapse compartment types, but the number of synaptic contacts differs substantially between layers. FPL is a law about structural roles, not connection magnitudes.
Compare to C. elegans F12: Pearson r=0.623. The nematode's smaller connectome (302 neurons) shows tighter linear correlation. In the much larger Drosophila brain (2,952 neurons), rank-order persistence (Spearman 0.663) remains strong while linear-scale correlation weakens (Pearson 0.363).
600 Million Years
The divergence between C. elegans (phylum Nematoda) and Drosophila (phylum Arthropoda) is approximately 600 million years. Different body plans, different architectures, different neuron counts (302 vs 2,952). Yet the same structural regularity holds in both. What is conserved is the structural principle: whichever neurons become hubs in one synapse compartment type also tend to be hubs in the other.
Layer Specialization
747 neurons (25.3%) are layer-divergent — high rank in one compartment type, low in the other. These are candidates for compartment-specific signalling (axo-axonic inhibitors, heterosynaptic modulators). The existence of 25% divergent neurons does not contradict the global FPL confirmation: FPL is a statistical tendency, not a deterministic rule.
Updated Scorecard
F13 adds 3 hypothesis tests: h1 CONFIRMED, h2 PARTIAL, h3 CONFIRMED.
New totals: 25 CONFIRMED, 4 DENIED, 2 PARTIAL = 31 pre-registered hypothesis tests.
Binomial probability of 25/31 confirmations by chance: p = 942,649 / 2,147,483,648 = 0.000439 (p < 0.001).
The law spans eight scientific fields: molecular biology, neuroscience, computer systems, ecology, linguistics/AI, computational geometry, particle physics, and medicine/epistemology.
Pre-registration: https://github.com/vladi160/preregistrations/blob/main/experiments/F13_drosophila_larval_v1.json Data: Winding et al. 2023, Science 379, eadd9330. DOI: 10.1126/science.add9330