Open Targets

@opentargets.org

708 Followers 38 Following

Public-private partnership using human genetics and genomics data for systematic drug target identification and prioritisation. http://blog.opentargets.org

Image caption: Overview of the VIDRA framework. (a) Data inputs. VIDRA integrates five complementary sources of genetic evidence: expression quantitative trait loci (eQTLs), protein quantitative trait loci, coding GWAS variants, AstraZeneca PheWAS variants, and ClinVar variants, together with gene-level rare variant burden tests. (b) Variant effect estimation. Each variant contributes a point to a dose-response model, where y represents disease effect and x represents either QTL effect size or a protein function score. Gene-level burden tests provide anchoring information for loss-of-function effects. Evidence is integrated using a hierarchical Bayesian model to estimate a gene–phenotype slope. (c) Outputs. For each gene–phenotype pair, VIDRA produces a posterior slope estimate, a posterior probability quantifying confidence in a non-zero slope, a Therapeutic Potential Score derived from a machine-learning classifier trained on approved drug targets, and a predicted direction of therapeutic modulation inferred from the slope sign. (d) Interpretation of slope direction. The VIDRA slope captures the relationship between protein activity and disease risk across the spectrum from loss-of-function (LoF) to gain-of-function (GoF). Positive slopes indicate that increased protein activity is associated with increased disease risk, whereas negative slopes indicate that increased protein activity is protective. Directionality is inferred from the combined genetic evidence and does not require explicit GoF or LoF annotation.
A man gestures at a scientific poster on a wall of posters, pointing something out to the woman standing next to him.
Next to a photo of Muzlifah Haniffa, Deputy director and head of the cellular genomics prorgamme at the Wellcome Sanger Institute, a quote reads: working through Open Targets has allowed us to partner with exactly the right people, without constraints. This kind of collaboration creates a real sense that we can accelerate progress towards clinical translation—moving faster and more effectively thanks to the breadth of expertise from many different disciplines.
22% of active Open Targets projects involve collaborations between EMBL-EBI and the Wellcome Sanger Institute. The accompanying graphic shows shaking hands as if making a business deal, with one hand in EBI green and one in Sanger blue.
Under a banner with the ESHG 2026 graphic, a list of presenters, title sessions, and timings. Gosia Trynka, Chair, S01. Hunting for effector genes from GWAS studies, June 13, 8:30 Chair, C17. In QTLs We Trust: Linking Genotype to Function, June 14, 10:30 Chair, S11. High-throughput perturbation methods to study disease, June 14, 14:00 Chair, S12. Blood as a window: genetic insights into health, aging, and immune function, June 15, 08:30 Tobi Alegbe, Chair, C17. In QTLs We Trust: Linking Genotype to Function, June 14, 10:30
Figure 2 from the publication: Heatmap of ΔPSI values for all 585 most completely characterized exons (>50% designed variants recovered), ordered by wild-type (WT) PSI (left color bar). Underneath is the reference to the paper, between the bioRxiv and Open Targets logos: Quarantani, G., et al. (2026) OpenSplice: the impact of half a million mutations on the alternative splicing of 600 human exons
Pleiotropy map of disease-associated genes arranged by chromosome around the outer ring, showing the number of unique diseases and therapy areas associated with the gene.
Figure describing self-supervised learning of gene programs using Tripso. The first section details the base model for gene representation, then gene programme representation, and finally a global model for cell representation. Tokenized scRNA-seq count data are used as input to a transformer model, which produces contextualized gene embeddings at single-cell resolution. Curated GPs are initialized from public databases. GP-specific transformers learn latent representations for each GP based on their constituent genes. A global cell block integrates GP representations to learn a unified cell-level embedding. Optionally, an additional GP discovery transformer is defined using highly variable genes or a user-specified gene set to infer data-driven GPs. The next section describes applications and downstream tasks, for example GP-centric representation learning, optimal transport in GP latent space to map populations across conditions, Gene-GP importance analysis, GP importance scores across groups, and identifying novel condition-specific GP. Conditions may represent health and disease, in vivo and in vitro, cell types, age groups...
26.03 Release highlights from the Open Targets Platform. Not covered by the next photos: update to the official Open Targets MCP.
Screenshot of the HSPC perturbation atlas, showing genes clustered by gene programme. Clicking on one brings up predicted perturbation results for that gene.
Map of regulatory cascades as determined with TAP-seq and Perturb-seq. The top row shows 85 enhancer-like CREs with significant target genes, the second row shows TAP-seq tested genes, and the third and fourth rows show genes involved at Perturb-seq. Lines between the layers represent significant connections.
Locus-to-Gene widget in the Open Targets Platform showing the Shapley values for features contributing to the prioritisation of two genes in a credible set.
A poker player lifts the corner of their cards on a table strewn with chips and playing cards, revealing a pair of kings.
Where to access Open Targets Platform data? Web interface, API and API playgrounds, Data downloads, Google BigQuery public dataset, Microsoft Azure Open Dataset, and new: AWS Open Data
Perturbation Catalogue home page showing key metrics
A brightly lit Christmas tree at night, with a building in the background. Pathways leading to the building are illuminated with lights dotted on each side.
Graphic with a 10 surrounded by fireworks and the Open Targets Platform logo
Quote accompanying a photo of Ellen McDonagh, Open Targets Translational Informatics Director and EMBL-EBI co-lead for the Open Psychiatry Project: The Open Psychiatry Project will translate the power of omics data into identifying potential new treatments for mental health conditions. This is the first major initiative to bring together mental health data for multiple stakeholders in an accessible way, and analysing this information collectively is a key step towards finding new targets for drug development.
A large group of people sat on steps in a building, smiling at the camera. The front row is holding bunting with the Open Targets logo.
People are sat in groups, working at laptops arranged. They are engaged in individual and small group work. The room looks busy, with whiteboards covered in writing along the back wall.
An Indiana-Jones style character finds a DNA print in the sand, and follows it until he finds a stone with glyphs. Brushing away the sand on the stone reveals the text: 10th anniversary Open Targets Hackathon, 21-22 October. He turns to the camera and says: "Are you ready for the hackathon?"
A man stands beside a poster board labelled 1014
Find us at ASHG 2025. The graphic lists speakers in the Featured Symposium, beyond transcriptomics: can imaging bridge the gap in variant interpretation, and poster presenters. The two speakers are Gosia Trynka and Helena Kilpinen. The four poster presenters are: Catherine Sutherland, Haerin Jang, Annalisa Buniello, and Daniel Considine. Four Open
Closing job ad for a postdoc in CRISPR meta-analytics and AI, from Open Targets and the Human Technopole. The role is in the Iorio Lab, salary up to 40k euros, apply by 5 October. The description reads: The Perturbation Catalogue project aims to create an online repository of data from post-perturbational experiments. You will focus on a meta-analytics project, developing computational methods for the reproducibility assessment, harmonisation and analysis or inter-study CRISPR datasets, aggregating data and developing and using innovative AI-based methods for therapeutic target prioritisation pipelines.
A gif from House MD, in which Dr House says: I finally have a case of lupus!
An example of an intervals widget for a variant in the Open Targets Platform
Open Targets Hackathon: one week left to submit a project
Text reads: Open Targets hackathon, 21-22 October 2025, compute sponsored by Google Cloud
Carlos Cruz smiles at the camera from a bridge over a small river in a European town on a cloudy day.
Star Wars meme, conversation between Anakin and Padme. Padme says: Ani do you want to use Open Targets Genetics to identify drug targets?". Anakin replies, getting progressively more upset: "It's dead. The new Open Targets Platform has completely replaced it. And not just replaced it like-for-like, there's SO much new data. And they keep adding new data every 3 months — I can't keep up! I hate them!"
Distracted boyfriend meme. The distracted boyfriend is labelled "Target discovery scientist", his girlfriend is "Open Targets Genetics", and the woman he is looking at is "Merged Open Targets Platform"
Infinity brain-type meme with progressively happier otters. The levels read: browsing top hits on GWAS Catalog, Running your own post-GWAS analysis, Open Targets Genetics, and Merged Open Targets Platform
Meme: Jeremy Clarkson and Richard Hammond comparing two cars. In the first image, Jeremy points to one labelled "Open Targets Genetics" saying, This is brilliant. In the second image, he points to another labelled "Merged Open Targets Platform" saying, "But I like this".
Comic with four squares. Death, a skull in a black cloak, carrying a scythe, takes Open Targets Genetics by the hand, saying "It's time to go". Open Targets Genetics asks, "Was I a good platform?", and Death says "yes. They leave together, side by side.
GWAS colocalisation metrics for overlapping sets from GWAS studies for one of the credible sets linking ESR1 to osteoporosis
Screenshot of the publication in Science, titled: Diversity and scale, genetic architecture of 2068 traits in the VA Million Veteran Program
Osteoporosis associations page on the Open Targets Platform showing GWAS associations evidence for ESR1
Registrations open!
Graphical abstract for the publication. The top section illustrates the in vitro model of T cell exhaustion, in which naive CD8 T cells OT-I are chronically stimulated with their cognate antigen— OVA peptide. This is followed by temporal phenotypic characterisation. This model recapitulates many critical hallmarks of exhaustion, including expression of canonical surface markers—ROS, PD1, CD39, 2B4 and LAG3— impaired proliferation, reduced cytokine production, decreased cytotoxic granule release, and metabolic alterations. The second section shows the identification of a gene signature of exhaustion between in vitro, mouse, and human TILs. Two in vivo models validate these results and establish a gene signature shared by in vitro and in vivo exhausted states. Critically, this signature is observed in tumour infiltrating T cells from multiple human tumour types, validating the translational potential of this model for discovering therapies.
UMAP of the 9 populations and 86 cell types identified after quality control and clustering
Runthrough of the main features of the Open Targets Platform 25.06 release Including: credible sets from GWAS Catalog and Million Veteran program; a new associations view featuring target interactors from Reactome, IntAct, Signor, and String; a revamped data downloads page powered by croissant; molecular structure viewers using AlphaFold and AlphaMissense predictions; gene burden data from Broad Institute of MIT and Harvard CVDI Human Disease Portal; pharmacogenetics annotations from PharmGKB through the European Variation Archive
Text reads: Announcement, Open Targets Genetics will be deprecated on 9 July 2025
Open Targets hackathon, 21-22 October 2025, Submit a project
Submit your questions for the Open Targets Platform Q&A session, 13 May, 3-4pm (UK Time)
Graphic reads: Q&A session, 13 May, 3-4PM UK time. Open Targets Platform Q&A. Have a question for us? Join a live discussion with panellists from across our core teams. Registration required
Comic illustration of the T Cell Show, overlayed with a play button
Webinar: from variant to drug target, using the expanded Open Targets Platform to explore the genetics of target discovery. 23 April 3-4PM UK time
Graphic advertising a tutorial at ISMB/ECCB 2025 in Liverpool, titled: Genomic variant interpretation and prioritisation for clinical research. Text reads: Explore the variety of annotations and techniques available to assess human variation and the implications of variant effects on human health and disease. Next to this are six stacked logos: EMBL-EBI, Ensembl, Open Targets, PDBe, UniProt, and Wellcome Sanger Institute.
Three men — David Hulcoop, Minister Patrick Vallance, and Ewan Birney — listen with interest to a Helena Cornu, a woman who is talking animatedly. They are seen in profile, dressed formally. Other people can be seen in the background.
Five people dressed officially stand around a plaque on a stand, smiling at the camera; two men on the left and a man and two women on the right. The slide projected behind them features logos from UKRI, Wellcome, and EMBL-EBI. From left to right: Ewan Birney, David Hulcoop, Jo McEntyre, Patrick Vallance, and Janet Thornton.
A graphic advertising a webinar for the Open Targets Platform. The title of the webinar is: From variant to drug target: using the expanded Open Targets Platform to explore the genetics of target discovery. The webinar will run on 23 April, 3-4 pm (UK time). Participants can register at ebi.ac.uk/training
A man and a woman stand on either side of a large screen, holding a red curtain in front of it. They drop the curtain at the same time to reveal the Open Targets Platform home page. The man turns to smile at the camera while the woman opens her mouth in mock surprise.
Graphic reads: New, (senior) bioinformatician. Salary range: bioinformatician, £36,924-£43,867; senior bioinformatician: £44,905-£53,349 Apply by 25 March.
Five people stand in a row in front of a red wall, posing and smiling at the camera. From left to right: Gosia Trynka, Joseph Powell, Tom Fitzgerald, David Hulcoop and Ellen McDonagh
Graphic reads: Coming soon, variant, study, and credible set data. 25.03 release, with the Open Targets Platform logo. The illustration shows cartoon hands typing on a laptop, where the screen shows a variant page in the Open Targets Platform.
Figure from the publication: Pseudo-temporal ordering of patients captures disease progression and identifies MASLD gene signature. A. Pseudo-temporal ordering of patients based on transcriptomic data recapitulates disease progression based on individual phenotypes, i.e. steatosis, ballooning, inflammation and fibrosis, and based on the NAS and MASLD scores. B-C. Pseudotemporal ordering of two independent orthogonal datasets based on the 145 genes that are most predictive of the trajectory provides a linear and clear separation of the disease stages (EPoS dataset (B) and Gubra dataset (C)). D. Functional enrichment analysis results of these 145 genes using EnrichR, where the Combined Score (c) has been calculated as c = log(p_value) * z_score, with p_value representing the Fisher exact test outcome, and z-score the deviation from the expected rank.
Under a read 'Closing' banner, text reads: Postdoc: Neurodegenerative Disease Immune Response Modelling. Salary: £38,000 to £49,000. Apply by 2 March. Leverage recent advances in stem cells, single-cell sequencing and spatial transcriptomics as well as CRISPR, to drive innovation in disease-relevant cell models, develop assays to measure immune response in neurodegeneration, and apply them at scale to advance our understanding of these disorders.
Part two, showing the new data added to our target prioritisation view