# Fine-mapping strategy in a difficult meta-analysis setting

**URL:** <https://isgw-forum.colorado.edu/t/fine-mapping-strategy-in-a-difficult-meta-analysis-setting/893>\
**Category:** Day 8\
**Created:** [June 11, 2026, 3:00pm UTC](https://isgw-forum.colorado.edu/t/fine-mapping-strategy-in-a-difficult-meta-analysis-setting/893 "2026-06-11T15:00:55Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![jeongah.lee](https://isgw-forum.colorado.edu/user_avatar/isgw-forum.colorado.edu/jeongah.lee/32/624_2.png) [@jeongah.lee](https://isgw-forum.colorado.edu/u/jeongah.lee)\
**Post date:** [June 11, 2026, 3:00pm UTC](https://isgw-forum.colorado.edu/t/fine-mapping-strategy-in-a-difficult-meta-analysis-setting/893/1 "2026-06-11T15:00:55Z")

</div>

I would like to get advice on how to deal with the following scenario.

**Scenario**

- I have GWAS summary statistics from a large meta-analysis of mixed-ancestry cohorts (multi-ancestry meta-GWAS).

- I do _not_ have access to cohort-specific GWAS summary statistics.

- I also do _not_ have a suitable LD reference panel that matches the effective LD structure of the meta-analysis (e.g. no per-cohort LD, no per-ancestry LD, and no large external biobank with a similar ancestry mix).

**Question**

- What are realistic options for fine-mapping in this situation, if any?
- Is it acceptable to use an imperfect external LD panel (e.g. a partially mismatched LD or a weighted mix of ancestry-specific LD panels) as an approximation, or should I skip fine-mapping and focus on alternative prioritization strategies instead?

Thank you!  
Best regards,  
Jeongah