Open Targets Platform 23.09 has been released!

Open Targets Platform Sep 21, 2023

The latest release of the Platform — 23.09 — is now available at platform.opentargets.org. It comes with a completely new, dynamic associations view: 'Associations on the Fly'.

Key points

Explore Associations on the Fly (beta), our refreshed associations page with new facets and functionalities
Read natural language summaries of a publication's target-disease association evidence
Data updates: Probes&Drugs, String, and more

Key stats

Metric	Count
Targets	62,733
Diseases	25,209
Drugs	17,096
Evidence	16,232,046
Associations	7,922,844

Find additional metrics on the Open Targets Community.

Associations on the Fly (beta)

“Associations on the fly” is a revamp of the current Open Targets Platform association page with new facets and additional built-in functionalities.

This is a major step in our ongoing effort both to refresh the look and feel of the Platform and to enhance the Platform’s capabilities to help build therapeutic hypotheses.

The Associations on the Fly view will eventually replace the classic associations view, so please let us know what you think by commenting on this post, or commenting on our post in the Open Targets Community.

Screenshot of the Associations on the Fly view. — The Platform's new Associations on the Fly view for Parkinson Disease. The user can quickly see which data sources have evidence supporting the association between Parkinson disease and the prioritised targets, and modify the relative importance of each datasource in the overall association scoring.

Key features

Dynamically modify the relative importance of the different data sources ‘on the fly’
View the supporting data directly in the associations table, and quickly compare evidence for different associations
Filter by datasource and data type (note: these are OR filters)
Pin rows for easy comparison; share your list of pinned targets

Walkthrough of the new Associations on the Fly view

OpenAI Literature Summarisation tool

For data features that link to publications, the Platform now provides the option for users to ask for a natural language summary of the target-disease evidence presented in the publication (when the full-text article is available and free to re-use).

Using LangChain, we first prompt OpenAI's GPT3.5 Turbo model to summarise relevant portions of the text, which we then ask it to summarise with the following prompt: "Can you provide a concise summary about the relationship between [target] and [disease] according to this study?". The resulting text is presented to the user.

An example of the summarisation tool at work in the Europe PMC detail widget showing text-mining evidence associating PRKN with Parkinson disease.

We hope this feature will help users better understand the available bibliography and provide contextual evidence in an easy-to-read manner.

Data updates

This release also integrates regular updates from our data providers. Of note, we have increased the scope of our drug index to include chemical probes reported by Probes&Drugs, increasing the number of compounds in our drug index. We have also updated our molecular interactions to include String Database version 12.0.

Additionally, improvements to our pipeline have increased the number of manuscripts with grounded entities; our dataset features over 17 million manuscripts with grounded entities (target, disease, or drug).

Finally, you can now view the schema of the files available for download in the Open Targets Platform in our data downloads page.

Let us know what you think by joining the conversation on the Community!