# L2G Score in the New Platform: Causality or druggability

**URL:** <https://community.opentargets.org/t/l2g-score-in-the-new-platform-causality-or-druggability/1815>\
**Category:** Data issue\
**Created:** [3 June 2025 13:24 UTC](https://community.opentargets.org/t/l2g-score-in-the-new-platform-causality-or-druggability/1815 "2025-06-03T13:24:10Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![BushraDohai](https://avatars.discourse-cdn.com/v4/letter/b/aeb1de/32.png) [@BushraDohai](https://community.opentargets.org/u/BushraDohai)\
**Post date:** [3 June 2025 13:24 UTC](https://community.opentargets.org/t/l2g-score-in-the-new-platform-causality-or-druggability/1815/1 "2025-06-03T13:24:10Z")

</div>

Hello Open Targets team,

I’m working on gene causality analysis and have been using the Open Targets Genetics platform, specifically the L2G (locus-to-gene) score, as a measure of gene causality. In the new release on the new platform, I noticed that terminology has shifted to focus on “Targets” instead of “Genes.” I wanted to clarify whether the **L2G score** in the new release still represents **gene causality only** , or if it now also incorporates **druggability or therapeutic potential**.

I rely on the L2G score to identify **causal genes** , using a cutoff of 0.5 for my analyses, and I want to ensure that the score still has the same interpretation. Could you please confirm whether the L2G score remains a pure causality measure or if it now combines druggability as well?

Thank you for your support!

---

<div class="post-metadata">

**Author:** ![hcornu](https://dub1.discourse-cdn.com/flex017/user_avatar/community.opentargets.org/hcornu/32/716_2.png) [@hcornu](https://community.opentargets.org/u/hcornu)\
**Post date:** [3 June 2025 13:36 UTC](https://community.opentargets.org/t/l2g-score-in-the-new-platform-causality-or-druggability/1815/2 "2025-06-03T13:36:04Z")

</div>

Hi @BushraDohai,

The L2G score predicts the likelihood a gene is causal for a particular GWAS locus. The machine learning model uses 28 features divided into four categories: distance, colocalisation, variant effect, and other (gene count, protein gene count, and credible set confidence).

You can find out more in the documentation: [Locus-to-Gene (L2G) | Open Targets Platform Documentation](https://platform-docs.opentargets.org/gentropy/locus-to-gene-l2g#features)

The score still has the same interpretation, and does not incorporate druggability or therapeutic potential.

Let us know if you have further questions!

Helena

---

<div class="post-metadata">

**Author:** ![BushraDohai](https://avatars.discourse-cdn.com/v4/letter/b/aeb1de/32.png) [@BushraDohai](https://community.opentargets.org/u/BushraDohai)\
**Post date:** [3 June 2025 13:43 UTC](https://community.opentargets.org/t/l2g-score-in-the-new-platform-causality-or-druggability/1815/3 "2025-06-03T13:43:25Z")

</div>

Thanks Helena for your quick reply 🙂