From a clinical question to a validated hit
Select a stage to explore the screening approach. This is a conceptual workflow, not an experimental dataset.
Define the context
Start with a clinical question and a molecular context, such as KRAS or DNA-repair alterations. Clinical genomics helps nominate hypotheses; it does not establish a functional dependency.
Explore the published evidence
These linked studies provide clinical and genomic context for hypothesis generation. They do not report results from this screening workflow.
Automated real-world data integration improves cancer outcome prediction
MSK-CHORD integrates real-world clinical and genomic data across five cancer types, including pancreatic cancer, to study clinicogenomic relationships and cancer outcomes. This observational data resource supports hypothesis generation rather than demonstrating a functional screening dependency.
Overview
We deploy genetic and small-molecule screens to systematically interrogate pancreatic ductal adenocarcinoma (PDAC) biology at scale. The goal is to move beyond single-gene hypotheses and let unbiased functional data nominate the targets, combinations, and resistance mechanisms that matter clinically.
Approach
CRISPR-based functional genomics screens identify genetic dependencies and synthetic lethal interactions across PDAC models.
Small-molecule and combination screens map which therapeutic pairings overcome adaptive resistance, informing rational trial design.
Computational pipelines integrate screen hits with clinical genomics so candidate targets are prioritised by translational relevance.
Platforms & methods
Genome-wide and focused CRISPR knockout, CRISPRi, and CRISPRa libraries are screened across patient-derived organoids and cell lines to map context-specific dependencies.
Arrayed and pooled small-molecule libraries — including epigenetic, DDR, and RAS-pathway probes — are combined in synergy matrices to nominate rational pairings.
Hits are triaged through orthogonal validation, isogenic models, and dose-response confirmation before advancing to in-vivo and translational testing.
Why it matters
PDAC is notoriously resistant to monotherapy. Screening lets us pre-empt resistance and assemble combinations before they reach patients, feeding directly into our KRAS and DNA-damage-repair programs.
Open questions
Which dependencies are conserved across the molecular subtypes of PDAC, and which are private to a genetic context such as KRAS dosage or HRD status?
How do screen-nominated vulnerabilities shift under therapeutic pressure, and can we anticipate the adaptive escape routes that drive clinical relapse?