Data and code for: Cortico-cerebellar connectome gradient compression and dissociated molecular vulnerability programs in Parkinson's disease

# Group-level derived data — data dictionary All files are **group-level and non-identifiable** (no subject-level rows, no IDs). Region index is the 283-region atlas throughout (ROI 1–283 = Brainnetome cortex 1–210, CIT168 subcortex 211–252, SUIT cerebellum 253–283), MNI152 space. Subject-level and raw data are controlled-access (see the release's `_CONTROLLED_ACCESS_NOT_DEPOSITED.md`). | Folder | Files | Description | |---|---|---| | `atlas/` | `yeo7_283_assignment.csv`, `A283.csv`, `atlas283_centroids.csv`, `atlas283_distance.npy` | 283-ROI atlas: Yeo-7/compartment labels, region metadata, MNI centroids (mm), pairwise Euclidean distance (283×283). | | `gradient_maps/` | `t_map_adjusted_G{1,2,3}.csv` | Per-ROI HC-vs-PD group t (covariate-adjusted: group+age+sex+meanFD), col `t_adj`, plus per-ROI p/q. | | | `g3_sig_rois_by_direction.csv` | The 19 G3-affected ROIs with anatomy, effect sizes, pole. | | | `eigentest_results.csv`, `hcpd_g3_meanloading.csv`, `hc_mean_g3.npy`, `pd_mean_g3_OFF.npy`, `variance_explained.csv` | Group eigenspectrum tests and group-mean G3 loading maps. | | `spatial_null/` | `surrogates_t_adj_G{2,3}.npy`, `surrogates_G{1,2,3}.npy` | Variogram (BrainSMASH) surrogate t-maps, 2000×283, used for spatial-null p-values (`q_spatial`). | | | `spin_test_*.csv` | Spatial-null p/q tables (PET and gene tests). *(files retain the legacy "spin" name; the statistic is the variogram-based "spatial" null renamed in the text.)* | | `transcriptomics/` | `gsea_all_pathways*_adj.csv`, `target3_pathways_adj.csv`, `observed_27_per_gene_abagen_G2.csv`, `kamath_*`, `wang_*`, gene lists | Genome-wide pre-ranked GSEA tables, 27-gene panel per-gene results, Kamath/Wang SN cell-type set results + gene lists. | | | `allen_expression_abagen.csv` (81 MB) | AHBA expression, abagen pipeline, 283 ROI × 15,633 genes. **Regenerable** from public AHBA (human.brain-map.org) with abagen; included for convenience. | | `neurotransmitter/` | `pet_roi_maps.csv`, `spin_test_pet_results.csv` | 20 neuromaps PET receptor/transporter maps parcellated to 283 ROIs, and their spatial-null results. | | `network_enrichment/` | `yeo7_enrichment.csv`, `yeo7_assignment.csv`, `yeo7_summary.txt` | Direction-stratified Yeo-7/compartment enrichment of FDR-significant ROIs. | | `reviewer_response/` | `tissue_confound_*.csv`, GMV `comparison_vs_baseline.csv`, `g3_19roi_survival.csv`, `t_map_gmvadj_*.csv` | Tissue-class confound sensitivity (group-level r/NES/q) and gray-matter-volume-adjusted G3 t-maps. | External inputs are public and not re-deposited: AHBA (human.brain-map.org), neuromaps (github.com/netneurolab/neuromaps), Kamath 2022 (GEO GSE140231), Wang 2024 (GEO GSE246638).