Ethics & Responsible Use

GENARCH is an educational resource. All content is population-level and derived from publicly available, peer-reviewed data. The atlas synthesizes genetic epidemiology, environmental health, and molecular biology literature into an accessible knowledge graph for students, educators, researchers, and community members.

No content on this site constitutes medical advice, diagnosis, or treatment recommendation. The atlas does not generate individual risk predictions or clinical decision support. Users seeking medical guidance should consult qualified healthcare professionals.

All visualizations, scores, and summaries describe population-level patterns of genetic architecture and environmental effect modification. Risk shift charts show relative changes across exposure strata at the population level — they do not predict outcomes for any individual.

Polygenic risk score notes are provided for educational context only. GENARCH does not compute, display, or store individual PRS values. Strength scores (0–1) are composite indices for comparing evidence across associations, not measures of personal susceptibility.

GENARCH does not collect, store, or process any personally identifiable information (PII). There are no user accounts, logins, cookies tracking identity, or analytics tied to individuals.

No personal genetic data (VCF, 23andMe, AncestryDNA, or any genotype format) is accepted, uploaded, or processed. The atlas operates entirely on published, aggregate population data.

If analytics are used, they are privacy-preserving and aggregate-only (e.g., Vercel Web Analytics, which does not use cookies or track individual visitors).

GWAS discovery cohorts remain predominantly European-ancestry (~85–90% of participants). This limits the transferability of genetic associations, polygenic risk scores, and gene–environment interaction estimates to non-European populations. Effect sizes, allele frequencies, and linkage disequilibrium patterns vary across ancestries.

Every disease page includes a Population Equity Notes section documenting ancestry representation, transferability limitations, and known data gaps. Where multi-ancestry replication data exists (e.g., PAGE, TOPMed), it is cited alongside European-derived estimates.

GENARCH does not make claims about genetic differences between racial or ethnic groups. Population-level genetic architecture data reflects study design biases, not biological hierarchies. Race is a social construct; ancestry is a genetic concept — both are handled with care throughout the atlas.

The Community Module presents environmental and health data using neutral, non-stigmatizing language. No community, census tract, or region is labeled as “diseased,” “unhealthy,” or “at-risk” in absolute terms. Instead, we use terms such as “higher modeled burden” and “elevated exposure levels” relative to regional or national baselines.

All community-level estimates include uncertainty bounds and explicit limitations (ecological fallacy, data currency, spatial resolution). Socioeconomic data is framed as correlative context, not causal attribution, with explicit acknowledgment of structural determinants of health.

Mechanism briefs are hypothesis-driven syntheses, not validated causal models. They represent the current state of evidence and are subject to revision as new data emerges.

Confidence ratings (low / medium / high) reflect evidence convergence across curated sources — they do not represent certainty. A “high” confidence rating means multiple independent evidence types converge, not that the association is proven.

Strength scores are composite measures normalized to [0, 1] and should not be interpreted as absolute effect sizes. They are designed for relative comparison across associations within the atlas, not for clinical or regulatory decision-making.

All content is subject to the biases and limitations of the underlying data sources, including publication bias, cohort selection effects, and temporal gaps between data collection and atlas publication.