🧠

Neurodegeneration Risk Predictor

model Created: 2026-04-04 05:03:31 By: test Quality: 90% ✓ SciDEX ID: model-29ce54ef-040c-4831-97b6-4850faa315
🧮 Model Details Deep Learning
ARCHITECTURE
pytorch
METRICS
accuracy
0.87
f1
0.84
EVALUATION CONTEXT
Dataset:SEA-AD Snyder et al. 2022 (Allen Institute) — donor-level neurodegeneration severity labels
Benchmark:OT-AD Target Ranking v1
Training data:SEA-AD bulk RNA-seq + neuropathology scores (535 donors, 84 brain regions)
Related Entities
Alzheimer's disease
Metadata
model_familydeep_learning
frameworkpytorch
parameter_count1500
training_config{'epochs': 100, 'batch_size': 32}
evaluation_metrics{'accuracy': 0.87, 'f1': 0.84}
_origin{'type': 'internal', 'url': None, 'tracked_at': '2026-04-04T05:03:31.017793'}
evaluation_datasetSEA-AD Snyder et al. 2022 (Allen Institute) — donor-level neurodegeneration severity labels
training_dataSEA-AD bulk RNA-seq + neuropathology scores (535 donors, 84 brain regions)
benchmark_nameOT-AD Target Ranking v1
benchmark_idbench_ot_ad_target_ranking_v1
descriptionMulti-layer neural network predicting neurodegeneration risk from gene expression, neuropathology, and clinical features. Trained on the Seattle Alzheimer's Disease Brain Cell Atlas (SEA-AD) dataset.
Linked Artifacts (14)
related🗃Allen Brain SEA-AD Single Cell Dataset100%
evaluated_on📓Pathway Enrichment Analysis for AD Targets90%
supports🧪Epigenetic Memory Reprogramming for Alzheimer's Disease70%
related🧪GFAP-Positive Reactive Astrocyte Subtype Delineation60%
related🧪APOE Isoform Expression Across Glial Subtypes60%
related🧪ACSL4-Driven Ferroptotic Priming in Disease-Associated Micro60%
related🧪Astrocyte MCT1/MCT4 Ratio Disruption with Metabolic Uncoupli60%
mentions📊Pathway diagram for TFEB50%
mentions📊Pathway diagram for TFEB50%
mentions📊Pathway diagram for SST50%
mentions📊Pathway diagram for CHMP4B50%
mentions📊Pathway diagram for PVALB50%
mentions📊Pathway diagram for RELN50%
mentions📊Pathway diagram for FLOT150%