Analysis code and Neurodegeneration Metabolite Score weight matrix: A 9-Metabolite Bioenergetic Crisis Signature Across the Neurological Disease Spectrum

This repository contains the Neurodegeneration Metabolite Score (NMS) disease-agnostic weight matrix and Python analysis scripts supporting the study "A 9-Metabolite Bioenergetic Crisis Signature Across the Neurological Disease Spectrum: Evidence from 274,241 UK Biobank Participants." The NMS synthesizes 9 plasma metabolites into a single composite score using inverse-rank weights derived from UK Biobank Nightingale NMR metabolomic data (N=274,241). All analyses use exclusively summary-level data accessed via Figshare (DOI: 10.6084/m9.figshare.29390471); no individual-level UK Biobank data are included. Files:- nms_agnostic_weights.csv: Disease-agnostic NMS weights for 9 signature metabolites (creatinine w=0.148, acetone w=0.139, acetoacetate w=0.134, glucose w=0.118, Ile/Leu w=0.115, Lac/Pyr w=0.102, Gln/His w=0.097, Cit/Glc w=0.075, lactate w=0.072; presence-weighted mean, sum=1)- taskP1_aim2_permutation_null.py: Permutation null test (n=10,000)- taskP1_aim3_forward_mr_python.py: Forward Mendelian randomization (45 pairs)- taskP1_creatinine_batched.py: Batched MR for Creatinine (76 SNPs)- taskP1_aim4_nms_theoretical.py: NMS weight derivation- taskP1_threshold_sensitivity.py: 3x3 threshold sensitivity analysis- taskP1_bootstrap_stability.py: Bootstrap stability (n=1,000)- taskP1_internal_split_validation.py: Internal consistency analysis- taskP1_epilepsy_overlap.py: Epilepsy comparator overlap analysis- taskP1_util_snp_range.py: SNP instrument range utility