How to retrieve & compare study-specific GWAS variant lists (e.g., UKB vs FinnGen) for a phenotype?

General
, , ,
1

Hi all!

I am trying to compare GWAS variant associations for a specific disease/phenotype across different large-scale studies available through Open Targets. I’m researching replication/conservation of candidate variations across populations/studies.

My specific goal is to:

  1. Identify a target disease/phenotype (e.g., Asthma EFO:0000270).
  2. Retrieve a full list of GWAS variants associated with this disease specifically from UK Biobank studies represented in Open Targets. I need variants preferably not overly filtered in advance (e.g., P-value < 1e-5, if feasible). For each variant, I need details like chromosome, position, reference allele, alternate allele, and importantly, the study-specific P-value (and effect size/beta if available) from that UK Biobank study.
  3. Retrieve a second full list of GWAS variants associated with the same disease, meeting the same significance threshold, but specifically from FinnGen studies represented in Open Targets, again including variant details and the FinnGen-specific P-value and effect size.
  4. Compare these two study-specific lists to identify overlapping variants and analyze differences/similarities in their association statistics (P-values, effect sizes, etc’) between the two cohorts.

Key Constraint & Context:

My primary challenge is obtaining the full lists of variants passing a less stringent significance threshold per each specific study source, rather than just lead variants or variants pre-filtered into credible sets.

  • I’ve mostly only found locus-level information or credible set summaries, which didn’t provide the comprehensive, study-specific variant lists I need.
  • The current Genetics Portal is excellent for visualization and exploring loci, but I haven’t found a clear way to export the raw, complete list of significant variants filtered by specific study source (e.g., “all FinnGen variants for Asthma with P < 1e-4”). It’s also marked for deprecation.

Could you please advise on the recommended method to achieve this using current Open Targets resources? Specifically:

a) How can I export the full list of associated variants (e.g., P < 1e-4) for a specific phenotype, filtered to originate only from a particular study source like UK Biobank or FinnGen, including their study-specific statistics (P-value, Effect Size)? Is this possible via the Genetics Portal UI, the Platform, or does it require using the API?
b) Does Open Targets provide any specific tools, methods, or API endpoints designed for this kind of systematic cross-study source variant comparison (e.g., directly comparing the significant variant list from UKB study X vs. FinnGen study Y for the same trait)?
c) If this exact workflow isn’t natively supported, could you suggest alternative approaches using Open Targets data (perhaps via bulk downloads or specific API queries) or point towards other resources that might facilitate this goal?

Thank you!