Method Profile
Mendelian-randomization target scores are followed by EPHX2 perturbation in patient-derived neural models and separate observational drug analyses.
Prioritize Alzheimer’s disease targets and repurposing hypotheses across genetic, cellular and health-record evidence.
A 1,229-target genetic screen; human iPSC-derived neurons and cerebral organoids; separate electronic-health-data cohorts.
Parallel ancestry-specific genetic analyses, EPHX2 variant overexpression/prime editing, and four drug-exposure association models.
Genetic inputs
Brain expression/protein quantitative trait locus datasets and Alzheimer’s disease genome-wide association studies, with European- and African-ancestry analyses distinguished.
Human neural systems
Patient-derived iPSC forebrain neurons; wild-type EPHX2 and p.Arg287Gln overexpression models and edited EPHX2 genotypes; cerebral organoid comparison.
Observational drug analysis
111,680 patients with mild cognitive impairment and 210 commonly prescribed drugs, with selected analyses repeated in an independent claims database.
Use genetic direction to nominate inhibition or activation, then assess the candidate with a separate functional measurement. Keep cellular results, animal drug effects and adjusted health-record associations as distinct evidence.
Compare wild-type EPHX2 with p.Arg287Gln in patient-derived neural models and use prime editing as an additional genotype test.
Neuronal protein comparisons report three measurements; the organoid comparison contains 14 wild-type and 16 mutant organoids. The described neuronal line is from one patient with late-onset AD.
Wild-type versus p.Arg287Gln overexpression; edited A/A versus G/G EPHX2 genotypes.
p.Arg287Gln reduced phosphorylated tau in neurons (P = 0.003) and organoids (P = 0.044); prime-edited A/A neurons had less EPHX2 protein.
Compare four drug-exposure definitions with subsequent AD incidence, adjusting covariates and using propensity-score matching.
111,680 patients with mild cognitive impairment; 210 frequently prescribed drugs.
Exposed versus unexposed and higher versus lower exposure; trazodone also compared with escitalopram in independent claims data.
Twelve drugs were associated with lower AD risk across all four exposure models; the independent trazodone–escitalopram comparison gave HR 0.78.
Patient-derived human neurons and cerebral organoids expressing the protective EPHX2 p.Arg287Gln variant had lower phosphorylated-tau measurements than wild-type models. The genetic target study also examined an EPHX2 inhibitor in mice and drug associations in health records.
