xQTL Analysis Workflow Builder

xQTL Analysis Workflow Builder#

Not set up yet? Start with Environment Setup to install the software stack and run a first pipeline on the example data shipped in this repository. Each module below links to its own page, where the inputs, expected outputs and full command interface are documented.

xQTL Analysis Workflow Builder

xQTL studies ask how inherited genetic variation changes a molecular phenotype—such as gene expression, splicing, methylation or polyadenylation—and whether that regulatory effect helps explain variation in a complex trait. The map follows that scientific argument from a measured molecular phenotype to a localized regulatory signal and, when appropriate, to evidence connecting the signal with disease biology. Genetic regulation is often tissue- and cell-context-specific, so the relevant path depends on what was measured and where it was measured (GTEx Consortium, 2020).

A practical map of the analysis

From biological samples to interpretable xQTL discoveries

The pipeline turns a broad biological question—does genetic variation alter a molecular process?—into progressively more specific evidence. Not every project needs every stage: the route should match the molecular phenotype, study design and claim being tested.

  1. 1Define the molecular phenotype

    Turn the assay into biologically interpretable features—expression levels, splice usage, methylation states or polyadenylation usage—in the tissue or cell population relevant to the question. Quality control protects the biological comparison from sample swaps, outliers and poorly measured features.

  2. 2Separate inherited regulation from other variation

    Individuals differ because of ancestry, environment, cell composition and technical effects as well as genotype. Accounting for these sources of variation makes the allelic effect easier to interpret as genetic regulation of the molecular phenotype.

  3. 3Map and localize regulatory effects

    Association testing identifies loci where genotype tracks the molecular phenotype. Multi-context analysis asks whether an effect is shared across tissues or cell types; fine-mapping narrows correlated variants to credible sets that quantify uncertainty rather than declaring a single causal variant (Wang et al., 2020).

  4. 4Relate regulation to complex traits

    Colocalization evaluates whether molecular and trait associations are consistent with a shared causal signal (Giambartolomei et al., 2014), while TWAS tests whether genetically predicted molecular levels are associated with the trait (Gusev et al., 2016). Together with functional enrichment, these analyses prioritize mechanisms but do not by themselves prove mediation or causality.

How to use this page: describe your data and goal, follow the highlighted route, open a module to inspect its real fixture files, and run the displayed sos run command.

Tell us about your data and goal

Inputs

Reference Data

Quantification

Molecular Phenotype Quantification

Bulk RNA-seq
Single-nuclei / pseudobulk
Alternative splicing
DNA methylation
Polyadenylation

Pre-processing

Data Pre-processing

Genotype
Phenotype
Covariate

Discovery

QTL Association Testing

Multivariate modelling

Cross-cohort Meta-analysis

Multivariate Mixture (MASH)

Regression

High-dimensional Regression

Individual level
Summary statistics level

Integration

GWAS Integration

Rare-variant xQTL

Interpretation

Enrichment & Validation

xQTL Modifier Score