Related Experiment Video
Updated: Oct 16, 2026
Biomarker Identification for Gender Specificity of Alzheimer's Disease Based on the Glial Transcriptome Profiles
Published on: May 20, 2024
Glial-adjusted residual transcriptomics with leave-one-out transcriptome-wide association study influence analysis
1Independent Researcher, Hong Kong, China.
Background:
Bulk transcriptomic studies of Alzheimer's disease are difficult to interpret because the diseased brain is not only transcriptionally altered, but structurally remodeled. Neuronal and synaptic loss, reactive gliosis, and vascular or extracellular-matrix changes can all shift measured expression.
Objective:
To reanalyze GSE5281 with glial marker-derived adjustment and integrate residual gene sets with Alzheimer's disease S-PrediXcan/transcriptome-wide association study (TWAS) results to prioritize genes relevant to disease pathogenesis and target discovery.
Methods:
We used age, composite glial score, and surrogate-variable adjustment to define residual upregulated and downregulated genes, then evaluated these sets in brain-tissue TWAS outputs using Stouffer meta-Z statistics, permutation/bootstrap set-level testing, and gene-level leave-one-out influence analysis.
Results:
The residual sets did not show significant set-level enrichment. AD_RESIDUAL_DOWN showed Stouffer Z = 0.862, permutation p = 0.6982, and bootstrap 95% CI = -3.18 to 5.38. AD_RESIDUAL_UP showed Stouffer Z = -1.564, permutation p = 0.5002, and bootstrap 95% CI = -5.43 to 2.26. However, gene-level analysis identified 41 influential gene-disease entries. GALNT6 was the strongest signal in the residual upregulated set and the strongest overall gene-level signal. C1QTNF4 and NTSR2 were the largest influence signals in the residual downregulated set, while EPHX2 ranked among the influential genes.
Conclusions:
These findings support glial-adjusted residual transcriptomics with leave-one-out TWAS influence analysis as a computational prioritization strategy for Alzheimer's disease, rather than as evidence of broad pathway-level genetic enrichment.
