A provenance-aware computational framework for cross-cohort transcriptomic prioritization in early-onset Alzheimer's disease
This repository contains analysis code and reproducibility materials associated with the study:
A provenance-aware computational framework for cross-cohort transcriptomic prioritization in early-onset Alzheimer's disease
Current release: eoad-computational-framework-v1.1.0
The repository includes workflows for:
- cross-cohort transcriptomic analysis;
- provenance-aware chemical-protein annotation;
- expression- and variance-matched null benchmarking;
- null-model ablation;
- K = 50-500 matching-neighborhood sensitivity;
- matching-balance diagnostics;
- cell-composition sensitivity;
- single-nucleus qualification;
- genetic and pharmacological analyses;
- numerical reproduction gates.
The frozen 913-gene EOAD candidate definition and the primary K = 250 matched-null specification were not redefined in version 1.1.0.
Historical filenames containing Biology_Direct, where retained within archived scripts, are preserved only for backward compatibility and do not denote a journal-specific analytical procedure.
Analysis and figure-regeneration code is released under the MIT License.