How can I join VISTA?
New Partners are welcome! The financial support increases the number of drugs that can be screened, expanding the potential for discovering therapeutic options and increasing the value of the dataset. Partners benefit from observing first-hand the experimental and data analysis techniques and troubleshooting, for potential future internal implementation. The project will also yield disease-related phenotypes and drug-disease hypotheses of interest to partners, as well as a machine-learning-ready dataset of disease-relevant phenotypes.
Non-profit groups can join as collaborators, bringing lab experimentation or data analysis capabilities.
Contact Anne Carpenter at the Broad Institute for more information (imagingadmin@broadinstitute.org).
What is the scientific foundation for VISTA?
The VISTA Consortium builds on a large-scale effort to functionally characterize pathogenic human coding variants using high-content imaging. By tagging, overexpressing, and imaging thousands of disease-associated variants alongside their wild-type counterparts, we previously identified ~250 disease-associated protein mislocalization phenotypes across ~100 genes (Lacoste, Haghighi, et al. Cell 2024). These phenotypes can be screened to identify drugs that rescue them, yielding correct localization and hopefully improved patient symptoms.
Rare diseases collectively affect hundreds of millions of people worldwide, yet only a small fraction have approved treatments. Traditional drug discovery requires custom assay development followed by compound screening and is typically conducted one disease at a time, requiring significant time and financial investment. VISTA aims to overcome this bottleneck by enabling parallelized functional testing of many variants and many drugs simultaneously.
What is VISTA’s experimental strategy?
VISTA’s underlying technology is Pooled optical image-based profiling. Hundreds of barcoded variants known to produce mislocalization phenotypes are combined in a single experiment. Each well contains pooled healthy and disease-associated alleles treated with one drug. Cells are imaged to measure protein localization and broader cellular morphology using high-content microscopy. Cell barcodes are read directly within the same cells using in situ sequencing to link each imaged phenotype back to the specific genetic variant expressed in that cell.
This pooled optical workflow makes it possible to test hundreds of disease variants against hundreds of compounds in parallel, dramatically reducing cost and increasing throughput compared to traditional approaches.
Libraries and Scale
The current VISTA effort includes:
- 350 ORF constructs representing 129 genes and nearly 200 unique variants with reproducible mislocalization phenotypes
- Several hundred compounds, including FDA-approved drugs and curated literature compounds targeting relevant pathways (such as unfolded protein response)
At scale, the consortium aims to test thousands of drugs across hundreds of diseases using a shared infrastructure. Ongoing optimization efforts include improving barcode detection efficiency, balancing representation of pooled constructs, optimizing cell distribution, and refining imaging conditions to maximize signal quality and reproducibility
Data Generation and Impact
Each experiment generates high-dimensional, single-cell image profiles linking:
- Genetic variant
- Cellular phenotype
- Drug treatment
- Phenotypic rescue or non-response
These datasets enable systematically identifying drugs that reverse disease-associated phenotypes. Promising hits will be subsequently validated in arrayed (non-pooled) confirmation screens, dose-response experiments, and additional cell types relevant to the disease context.
Beyond drug repurposing, the dataset and platform can be extended to:
- Identify novel targets for diseases (using genetic perturbations rather than drugs)
- Group diseases with similar mechanisms, incentivizing commercial drug development
- Map functional relationships between proteins and drugs
- Inform mechanism-of-action studies
- Support downstream translational and industry development efforts
Consortium Structure and Governance
VISTA operates as a structured public-private consortium. The consortium model enables:
- Shared scientific input
- Coordinated experimental prioritization
- Scalable data production
- Collaborative review of findings
Partners may nominate diseases, genes, variants, or compounds of interest, and consortium-generated data will be shared publicly following review and analysis, enabling broad scientific impact.
Long-Term Vision
The long-term goal of VISTA is to transform how rare and genetically defined diseases are approached, moving from bespoke, disease-specific screening to a scalable, parallelized therapeutic discovery platform. By systematically mapping variant-driven cellular phenotypes and identifying compounds that restore healthy cellular states, VISTA aims to reduce the time, cost, and uncertainty associated with rare disease drug discovery.
Ultimately, we aim to test thousands of approved drugs and nutraceuticals for thousands of diseases.