Abstract "This study investigates whether protein backbone conformations exhibit reproducible alignment to specific mathematical targets, specifically the Golden Ratio (ϕ) and π. Using a systematic panel of 2, 400 protein structures (comprising over 534, 000 ϕ/ψ angle pairs), we tested two preregistered geometric heuristics: golden-ratio proximity and toroidal distance enrichment near π. Applying a 'protein-equal weighting' methodology to prevent large structures from dominating the results, the data was split into discovery and validation subsets. Both heuristics replicated across splits with high statistical significance (p<0. 0001). The results suggest that protein folding is not merely a stochastic physicochemical process but follows a highly optimized, non-random geometric framework, potentially functioning as a biological form of 'data compression' for structural stability. " Description This goes in the "Notes" or "Description" box to explain your methodology. "Key Findings: Golden Ratio (ϕ) Proximity: Observed mean deviation of ~0. 779 vs a null mean of ~0. 838, indicating proteins 'tune' their angles toward the golden ratio for stability. π-Scale Enrichment: A consistent ~12. 6% hit rate for π-correlated geometry, significantly higher than random null models (p<0. 0001). Scale: This analysis represents a high-power validation using a non-redundant dataset of 2, 400 proteins. Methodological Perspective: The identification and testing of these specific mathematical constants were driven by a neurodivergent research framework (ASD/ADHD). This approach prioritized cross-disciplinary pattern recognition—bridging number theory and molecular biology—to identify structural 'signatures' that are often overlooked in traditional biochemical analysis. Files Included: proteingeometryₘanuscriptdraftᵥ6. pdf: Full technical report and discussion. perₚroteinₛummaryᵥ6. csv: Raw data for all 2, 400 proteins analyzed. paperᵣeadyᵥalidationfigureᵥ6. png: Visual distribution of the angular clustering. " proteingeometryₐnalysisᵥ6ᵥalidationfixed. py: Code for testing
Malin Hess (Wed,) studied this question.