Reframe 6,048 family clusters from QC artifacts into connected genomic systems with typed architecture: Subjects, IBDGraph, PRSField, PhenotypeField, AncestryContext, ReliabilityProfile, AdmissibilityState, ProvenanceChain

Per-cluster GWAS inflation quantification — not global pruning but per-cluster risk scoring. Completely different analytical capability from standard relatedness removal

141 suspect duplicate pairs and 6,048 family clusters already identified — FamilyClusterObject registration can begin immediately from existing IBD output files

DI-GOG: register first real edge (IBD pair) to promote DI-GOG from CANDIDATE-STUB to CANDIDATE, unblocking PO-GOG-001

IBD family clusters are by construction ancestry-homogeneous — naturally stratified anchors for PRLS_t calibration. A fully EUR-homogeneous cluster with high PRLS_t on PGS000138 provides empirical calibration not possible at individual level

GR_002 already admits PGS000138 (Wray et al. 2018). Depression PRS computed within GR_002 clusters demonstrates PRS propagation fields concept — immediate low-cost deliverable

eMERGE Phase IV GIRA returns cross-ancestry PRS for 11 conditions including depression, CHD, T2D. DI-PRS PRLS_t governance provides admissibility scaffolding for all 10 returned PRS scores

Ancestry tensor A_j (PO-PRS-004) can be partially addressed via eMERGE k-means ancestry clusters: 17,246 AFR · 3,720 ASN · 84,142 EUR already classified

Standard approach: remove relatives, continue. DI approach: quantify GWAS inflation risk per cluster — completely different analytical capability

Marshfield anomalous IBD cluster (108/139 samples from family study) is a real example — quantifying its inflation contribution is a concrete DI-GWAS demonstration

Herpes zoster GWAS (λ ~1.02) and CAAD GWAS both use IBD pruning. DI can provide admissibility claims with per-cluster leakage accounting

~1,800 ICD-9 PheWAS codes available. Depression + dementia phenotypes mapped to IBD family clusters is pre-analytic discovery of shared architecture without running TWAS/PWAS

Dynamic phenotype fields (CAP 07): IBD + PRS + Phenotype composition. Where PRS exists without phenotype = likely investigation zones for missing heritability

eMERGE CAAD PheWAS (Palmer et al.): rs6952610 PheWAS showed "cerebral ischemia" as top result. DI can provide admissibility chain for downstream PheWAS claims

eMERGE IV returns PRS for 11 conditions to 25,000 diverse individuals via GIRA report. DI-PRS PRLS_t provides admissibility framework for all 10 PRS returns

GIRA risk hierarchy: Monogenic > Polygenic > Family history. DI-GOG provenance chain maps cleanly to this three-tier admissibility structure — natural architectural alignment

CRC PRS excluded from GIRA because it didn't validate in Hispanic or African ancestry. DI-PRS PRLS_t would have flagged this at the score admissibility gate before return

27 of 83 batches excluded (Rsq < 0.3). DI-GENOME L3 (Imputation R²) formalizes this as a subject-level quality dimension within each FamilyClusterObject — not a global filter

Batch Rsq is highly predictable: log(sample count) + log(variant count) explain ~80% of variance. DI can formalize this as a structural reliability predictor for GOL state assignments

Mean Rsq > 0.9 for common variants (MAF > 0.05) but drops below 0.3 in 0.00001–0.0001 MAF bin. DI rare-variant admissibility (DGV_rare) maps directly to this distribution

Xian et al. eMERGE GWAS: PheKB 1095 (depression) shows rg = 0.83 vs MDD2019. Novel HLA/MHC-II and IGHV adaptive immunity loci identified. IBD family clusters are natural hypothesis generation units for depression–dementia co-architecture

Wingo et al. dementia susceptibility work at Tulane: shared genomic architecture between depression and early dementia is exactly what DI cluster phenotype mapping would surface as pre-analytic signal

Cross-sector IDR coupling (β_GC) required before GR_003 composite claims are admissible. IBD cluster stratification as design variable provides an anchor for β_GC empirical estimation

Tc empirical calibration for all DI-GENOME stub modules

β_GC cross-sector IDR coupling (G↔C sectors)

Ancestry calibration tensor A_j for PRLS_t

First real GOG edge registration (PO-GOG-001)

PO-GENOME-007: PRLS_Cluster functional form definition

PO-GENOME-008: I_residual_local(cluster) formalization

PO-GENOME-009: FamilyClusterObject architecture definition

ODD-GENOME-001: FamilyClusterObject classification

Map depression + dementia phenotypes to 6,048 eMERGE family clusters — T1 observation, existing data, no new computation

Compute PRS within GR_002-admitted clusters using PGS000138 — demonstrates family PRS propagation field concept

Identify EUR-homogeneous family clusters for PRLS_t calibration — highest-probability path to partial Tc closure

Design GR_003 with IBD structure as stratification variable — depression PRS → dementia outcome cross-trait admissibility