GWAS of plasma GFAP in East Asian cohort

PRIMARY OUTCOME

plasma GFAP levels

EXPECTED OUTCOMES

1. **Primary signal at APOE locus:** rs429358 associated with plasma GFAP at p = 1.2×10⁻¹², ε4 carriers showing 1.4-fold higher GFAP vs ε3/ε3 (95% CI: 1.28-1.54) 2. **Novel signal at SLC10A7 locus:** rs1145016 (missense, Ala91Thr) associated at p = 3.7×10⁻⁸, T allele associated with 1.18-fold lower GFAP (95% CI: 1.10-1.27) 3. **Heritability estimate:** SNP-based heritability (h²SNP) of plasma GFAP = 0.22 (SE 0.04), consistent with prior European studies despite population differences 4. **Genetic correlation:** rG = 0.65 (SE 0.12) between plasma GFAP and AD risk, supporting shared genetic architecture 5. **eQTL colocalization:** SLC10A7 rs1145016 colocalizes with eQTL signals in brain tissue (PP4 = 0.78), suggesting cis-regulatory mechanism 6. **Replication power:** >90% power to detect effect size of 0.15 SD at α = 0.05 in 1,500-person BioBank Japan replication cohort 7. **Cellular validation:** siRNA knockdown of SLC10A7 in iPSC-astrocytes reduces secreted GFAP by 1.32-fold (95% CI: 1.18-1.48) vs scrambled control ---

SUCCESS CRITERIA

- **Primary GWAS threshold:** At least one SNP at APOE or SLC10A7 reaches p < 5×10⁻⁸ in discovery cohort (n=2,000); if APOE signal fails, task is considered failed regardless of other signals - **Effect size magnitude:** Detected SNP effects must exceed 0.12 SD change in plasma GFAP per allele; smaller effects (<0.10 SD) indicate insufficient power or population stratification - **Replication requirement:** Lead SNPs must replicate with p < 0.05/3 = 0.017 (Bonferroni for 3 primary hypotheses) in BioBank Japan cohort with directionally consistent effect - **Genomic inflation control:** λGC must be 0.98-1.05; λGC > 1.10 indicates systematic inflation and requires re-analysis with expanded covariates - **eQTL validation:** SLC10A7 risk genotype must show concordant eQTL effect in GTEx brain tissues (p < 0.05) to confirm biological mechanism - **Astrocyte assay:** siRNA knockdown must produce >30% reduction in SLC10A7 mRNA (qPCR) with corresponding GFAP protein change; failure indicates locus is not functionally relevant - **AD genetic overlap:** Polygenic risk score for AD must predict plasma GFAP at p < 0.01 in fully adjusted model; absence of genetic correlation undermines biological interpretation

PROTOCOL

### Phase 0: Study Design & Population Recruitment (Weeks 1-8) **Timepoints:** Week 1-2 (design), Week 3-8 (recruitment) - **Cohort:** 2,000 East Asian participants from K-ROAD (Korean R&D database), aged 60-85 years, both sexes - **Stratification:** 1,000 AD cases (NINCDS-ADRDA criteria), 1,000 cognitively normal controls - **Inclusion:** East Asian ancestry (self-reported + PCA validation), available plasma, informed consent - **Exclusion:** Other neurological diseases, recent stroke (<6 months), autoimmune conditions ### Phase 1: Sample Collection & Processing (Weeks 6-12) **Timepoints:** Fasting blood collected at 8:00-10:00 AM - **Blood collection:** 10 mL EDTA plasma + 5 mL serum per participant - **Centrifugation:** 1,500 × g, 15 min, 4°C within 30 min of collection - **Aliquoting:** 4 × 250 μL plasma aliquots, 2 × 250 μL serum aliquots, stored at -80°C - **DNA extraction:** QIAamp DNA Blood Kit (Qiagen, cat# 51106) from buffy coat - **DNA QC:** NanoDrop 260/280 ratio >1.8, Qubit fluorometry >50 ng/μL ### Phase 2: Genotyping (Weeks 10-18) **Timepoints:** Week 10 (DNA QC), Week 12-16 (genotyping), Week 17-18 (QC) - **Platform:** Illumina Asian Screening Array (ASA) + custom Content (SLC10A7, APOE flanking regions) - **Genotyping:** 730,000 markers + 5,000 custom SNPs - **APOE genotyping:** rs429358 (C>T), rs7412 (T>C) for ε2/ε3/ε4 determination - **SLC10A7 SNP selection:** 1000 Genomes East Asian LD tagging (r² > 0.8) + known missense variants - **Quality control:** - Sample call rate >98%, SNP call rate >99% - HWE p > 1×10⁻⁶ - Remove ancestry outliers (>4 SD from East Asian cluster in PCA) ### Phase 3: Plasma GFAP Measurement (Weeks 8-20) **Timepoints:** Single measurement per participant at baseline - **Assay:** Simoa Human GFAP Advantage Kit (Quanterix, cat# 103346) - **Sample preparation:** Thaw plasma on ice, centrifuge 10,000 × g, 5 min, 4°C - **Dilution:** 1:4 in sample diluent - **Instrument:** Simoa HD-X Analyzer (Quanterix) - **Detection range:** 8.22-4,000 pg/mL - **Intra-assay CV:** <5%, Inter-assay CV: <10% - **Batch randomization:** Blind duplicates (10% of samples) randomly distributed across plates ### Phase 4: GWAS Analysis (Weeks 18-28) **Timepoints:** Week 18 (imputation), Week 20-24 (association), Week 25-28 (validation) - **Imputation:** Michigan Imputation Server (Eagle v2.4, minimac4), TOPMed r2>0.3 - **Association model:** Linear regression (quantitative trait) in PLINK 2.0 - **Covariates:** Age, sex, APOE ε4 status, PC1-PC10, batch effect - **Genomic inflation:** λGC < 1.05 - **Significance threshold:** Genome-wide (p < 5×10⁻⁸) + suggestive (p < 1×10⁻⁵) - **Conditional analysis:** GCTA-COJO for independent signals - **SLC10A7 expression QTL:** eQTL analysis in DICE database + Genotype-Tissue Expression (GTEx) v8 ### Phase 5: Replication & Functional Validation (Weeks 26-40) **Timepoints:** Week 26-30 (independent cohort), Week 30-36 (cell model), Week 36-40 (bioinformatics) - **Replication cohort:** 1,500 East Asian participants from BioBank Japan - **Cell model:** iPSC-derived astrocytes from 3 donors per genotype (SLC10A7 rs1145016 variants) - **siRNA knockdowns:** Lipofectamine RNAiMAX (Thermo Fisher), 30 nM siSLC10A7 (Santa Cruz, sc-90178) - **GFAP ELISA:** Cell supernatant collected at 72h post-transfection, R&D Systems Duoset (cat# DYC2468) - **Western blot:** Anti-GFAP (Cell Signaling, cat# 8245), anti-β-actin loading control ---

LINKED HYPOTHESES

HTML iframe

<iframe src="http://scidex.ai/artifact/exp-a6d16514-1a7a-45c3-8ca8-df341e213c0a?embed=1" width="100%" height="600" style="border:0;border-radius:8px"></iframe>

Markdown link

[GWAS of plasma GFAP in East Asian cohort](http://scidex.ai/artifact/exp-a6d16514-1a7a-45c3-8ca8-df341e213c0a)

Direct URL

http://scidex.ai/artifact/exp-a6d16514-1a7a-45c3-8ca8-df341e213c0a

Preview