Credible Sets

Our fine-mapping analyses for a study determine the credible sets. A credible set is a set of genetic variants near a genetic association signal that is predicted, with a specific probability, to include the causal variant for that signal. In the Platform, these are 95% credible sets; in other words the variants in the credible set cover the top 95% likelihood of containing the causal variant.

On the credible set page, you can look through:

• The list of variants in the credible set
  
• The list of likely causal genes from our Locus-to-Gene (L2G) analysis
  
• Colocalisation results of overlapping sets from GWAS or molecular QTL studies which, respectively, can indicate other complex traits that may be related to the one of interest, and the likely functional consequence of the credible set. 

Watch this video, from 23 April 2025, for an overview of our credible set pages:

Explore the credible set page for yourself: https://platform.opentargets.org/credible-set/4b11d5bd27a3c8a21328ecf2d08fd044

What is a credible set?

For a more in-depth explanation of credible sets, watch this video in which Daniel Considine responds to a question from the Q&A on 23 April 2025:

Does L2G always pick the closest gene?

Based on genetic and functional genomics traits, our Locus-to-Gene (L2G) machine-learning algorithm ranks the most likely causative genes at each GWAS credible set. The likelihood that a gene is causal for a particular GWAS locus is measured by the L2G score, ranging from 0 to 1.

One of the attendees of our webinar on 23 April 2025 asked whether there are good examples where L2G assigns a higher score to a gene that is not the nearest one. Daniel Considine replied: