Methods

Overview

This portal provides access to harmonized brain multi-omic xQTL results that can be explored by gene, variant, or genomic region. It integrates data from cohorts such as ROSMAP, MSBB, Knight-ADRC, and MiGA to support discovery, interpretation, and downstream analysis in one searchable resource.

The purpose of this page is to explain what data are included, how results are organized, and how to use the portal effectively. For users who want implementation-level detail, links to workflow, dataflow, consortium resources, and GitHub code can be provided throughout this page. Data sources and cohorts

The portal brings together harmonized xQTL results from major brain-focused cohorts and related resources, including ROSMAP, MSBB, Knight-ADRC, and MiGA. The current portal highlights 14 brain regions, 7 cell types, 6 omics types, more than 13,000 samples, and over 300 million xQTL associations. These datasets are organized to support cross-study exploration while preserving key context such as molecular phenotype, brain region, and cell type. See < XXX Dataflow diagram of this project> for dataflow. Links: XXX Detailed cohort descriptions.

Analysis pipeline

At a high level, the portal organizes results through 2 main stages: 1) preprocessing, and performing association testing; 2) annotation, harmonization and integration. With both stages, this consistent framework allows users to explore xQTL associations across studies and molecular layers with any ease

• Preprocessing — prepare the genotype data and molecular phenotype data per cohort.
• Association testing — identifying statistical associations between genetic variants and molecular traits.
• Annotation — linking results to genes, loci per genomic context
• Harmonization — standardizing data structures, identifiers, and context across cohorts.
• Integration — making results searchable, browsable, and downloadable through the portal.

View Detailed workflow here , together with affiliated GitHub code repositories < https://github.com/statfungen/xqtl-protocol/tree/main/code , HipFG URL>

Data model and terminology

The portal uses a consistent set of terms to help users interpret results across datasets and views.

• Omics type — the molecular layer being analyzed, such as expression, methylation, protein, histone acetylation, or splicing.
• Cell type — the cellular population in which the molecular trait was measured or inferred.
• Brain region — the anatomical source of the sample.
• Significant association — an xQTL result that passes the statistical threshold defined in the analysis framework.
• Locus — a genomic interval containing one or more variants or associations.
• Track — a visual layer in the genome browser used to inspect local genomic context.

How to use the portal

The portal supports four main entry points that are already reflected on the homepage: search by gene, search by variant, search by region, and browse tracks. Example queries are also currently shown on the front page of the site. A simple way to start off is:

1. Search for a gene to find associated molecular traits.
2. Search for a variant to inspect evidence for a specific signal.
3. Search for a region to review all xQTLs within a locus.
4. Open tracks to visualize local genomic context interactively

Example workflows

Find evidence for a gene

Enter a gene symbol such as APOE to review associated xQTL results across available datasets and molecular phenotypes. Then move to the track browser if you want to inspect the locus in more detail.

Inspect a variant of interest

Search for a variant such as rs7412 to examine associated molecular traits and relevant dataset context. This is a useful starting point for following up candidate GWAS signals.

Explore a genomic interval

Enter a region such as chr19:44908000-44909000 to retrieve all xQTL associations within that interval. This is especially helpful when a locus contains multiple genes or multiple candidate signals.

Browse tracks

Use the genome browser view to inspect xQTL tracks and local genomic structure. This helps users move from result lists to a more visual understanding of the locus.

Output interpretation

When interpreting any QTL results, users should first check:

• the molecular trait or target;
• the xQTL type;
• the study or dataset context;
• the brain region or cell type;
• the significance or effect information.

Common mistakes to avoid:

• treating association as proof of causality;
• ignoring tissue, region, or cell-type specificity;
• focusing on a single association without looking at the broader locus context.

Downloads and reproducibility

Access bulk xQTL summary statistics, metadata, file descriptions, and release information. Code repositories and workflow documentation support reproducibility, transparency, and reuse.

Data download options

Choose from three approaches based on your needs:

1. Search results. Download association records directly from search results under the "Associations" tab at xqtl.niagads.org.
2. Significance-based downloads. Retrieve all xQTLs or significant xQTLs via the NIAGADS data portal: dss.niagads.org/open-access-data-portal/ (Filter DATASET by NGxxxxx, then FILE NAME by xxx)
3. QTL type-based downloads. (Filter DATASET by NGxxxxx, then FILE NAME by xxx)

XXXMetadata files describing each xQTL file: xxx XXXFile format definitions: xxx

Code and workflow

• Summary stats generation: https://github.com/statfungen/xqtl-protocol/tree/main/code
• Summary stats QC by hipFG: xxx

Related resources

• Visit the NIAGADS xQTL collection for all FunGen-xQTL project resources beyond these summary statistics.
• Visit other pipelines developed by the FunGen-xQTL team: https://statfungen.github.io/xqtl-protocol/README.html
• Visit this for all the FunGen-AD consortium members: https://adsp-fgc.niagads.org/xqtl-resources/fungen-ad-members/

Help

If you are new to the portal, start with one of the example queries shown on the homepage and then move to track browsing or downloads depending on your goal. For data questions or technical issues, contact help@niagads.org