Genome-wide association studies (GWAS)
Association between variants and phenotype on your genotypes: quality control, population structure control and analysis with regenie. Results with multiple-testing correction, ready for publication or decision-making.
From genotype to finding
A reproducible pipeline on genotypes you already have (array or sequencing): we do not reanalyze the sequencing, we focus on the statistics of association and its interpretation.
Quality control
MAF, Hardy-Weinberg equilibrium and missingness filters per individual and variant (PLINK2). Only informative markers pass into the analysis.
PLINK2Population structure
Principal components (PCs) as covariates to avoid spurious associations driven by ancestry. Additional covariates (sex, age, batch) on request.
PCs · covariatesAssociation (regenie)
Ridge null model and association tests for quantitative or binary traits, with an allele-count threshold. Scales to large cohorts.
qt / bt · minMACResults and report
Results table (beta, SE, statistic, p) with top hits, multiple-testing correction and an interpretable summary.
TSV + reportCompatible formats
Genotypes in PLINK format (bed/bim/fam) or VCF; phenotype in a table (FID IID + trait + covariates).
PLINK · VCFCritical reading
We interpret the findings in context: effect size, replication and biases. Unreplicated results are hypotheses, not a diagnosis.
Responsible scientistFour steps
1. Data
You send us genotypes and phenotype; we define the trait, covariates and model.
2. QC + PCs
Quality filters and calculation of principal components.
3. Association
regenie (step 1/2) with the agreed covariates.
4. Report
Top hits, multiple-testing correction and an interpretable summary.
What it includes
QC (MAF/HWE/missingness), PCs, qt/bt association with covariates, results table, top hits with multiple-testing correction and an interpretable summary. Volume and academic discounts.
Honest limits
A GWAS is not a diagnosis: validity depends on controlling for structure and consanguinity, and findings require replication. We report with multiple-testing correction and deliver everything as research (RUO).
Other research services
Do you have genotypes waiting for a GWAS?
Tell us the cohort, trait and covariates; we get back to you with a quote and timeline within 48 h.