Teaching pack: Drug Discovery Platforms

Drug discovery platforms are the tools used to find the next drug: gene screens, organoids, models in mice and AI. The section spans CRISPR functional genomics including DepMap, patient-derived organoids and xenografts, ex vivo drug sensitivity testing, structure-based and AI-driven design, degrader platforms and conjugation chemistry. Its own technology list holds Chai-1 and Chai-2, RFdiffusion and ProteinMPNN from the Baker Lab, BioEmu, Phenom-2 and Recursion OS, and Chemistry42 from Insilico. Technologies that link back include CRISPR screens, patient-derived organoids and xenografts, functional drug testing, PDAC organoid pharmacotyping, BH3 profiling, whole-genome sequencing, single-cell and spatial profiling, proteomics, synthetic lethality approaches and site-specific conjugation.

• Adaptive therapy (evolution-based dosing): Instead of hitting a tumour as hard as possible, adaptive therapy gives just enough drug to keep it in check and stops when it shrinks, so drug-sensitive cells survive to compete with resistant ones. A pilot trial in prostate cancer roughly doubled the time to progression on abiraterone.
• Agent-based and multicellular simulations: Instead of equations for average behaviour, agent-based models simulate every cell as an individual with rules for dividing, moving, dying and signalling, producing virtual tumours in which immune attack, drug delivery and evolution can be watched and tested.
• AI compute and model platforms for oncology: AI compute platforms are the GPUs, model libraries, and cloud services that pathology, radiology, and drug-design AI run on.
• AI-driven drug & target discovery: Using machine learning to pick targets, design molecules and antibodies, and predict which ADC will work.
• AlphaFold 3: Predicts the 3D shape of proteins together with DNA, RNA, small molecules and antibodies, the starting point for much modern drug design.
• Angiogenesis and vascular normalisation models: Models of how tumours recruit blood vessels, and of Rakesh Jain's idea that anti-angiogenic drugs at the right dose normalise rather than destroy those vessels, improving drug and oxygen delivery for a window of days.
• Bacteriophage-based tumour delivery: Bacteriophage delivery uses viruses that infect bacteria, not human cells, as engineered shells whose coat proteins display tumour-homing peptides or antigens and carry drugs or vaccines. They are cheap and cannot replicate in people, but the work is preclinical: no oncology phage trial had reported efficacy by 2026, and the body clears them quickly.
• BH3 profiling (functional apoptosis testing): A lab test that measures how close a leukaemia cell is to self-destructing, and which survival protein is holding it back, to predict response to venetoclax-type drugs.
• Biobanking and tissue procurement: Freezers full of consented tumour samples with matched clinical data, which every biomarker and drug programme depends on.
• BioEmu (Microsoft): BioEmu is a Microsoft generative diffusion model that predicts the range of shapes a protein moves between, not one static structure, thousands of times faster than molecular dynamics simulation. For cancer drug discovery that can reveal transient pockets, as in KRAS, that static predictors miss, but its outputs are approximate and validated mainly on small proteins.

• Drug discovery roadmap: screening in mice → maps of dependency → designing in silico · Screening nature and hoping (historic)
• Drug discovery roadmap: screening in mice → maps of dependency → designing in silico · Targets, structures and high-throughput screens (historic)
• Drug discovery roadmap: screening in mice → maps of dependency → designing in silico · Maps of dependency and models closer to the patient (current)
• Drug discovery roadmap: screening in mice → maps of dependency → designing in silico · Structure prediction and generative design (current)
• Drug discovery roadmap: screening in mice → maps of dependency → designing in silico · New modalities as platforms (current)
• Drug discovery roadmap: screening in mice → maps of dependency → designing in silico · Functional precision medicine (emerging)
• Radical oncology: what could change the war by 2035 · Ideas already being tested against a control arm (current)
• Radical oncology: what could change the war by 2035 · Living drugs, logic gates, and designed proteins reach decision points (emerging)
• Radical oncology: what could change the war by 2035 · Radiation and radiopharmaceuticals get a second act (emerging)
• Radical oncology: what could change the war by 2035 · Monitoring becomes continuous and selection becomes spatial (emerging)
• Radical oncology: what could change the war by 2035 · Writing to the genome and the epigenome inside a tumour (speculative)
• Radical oncology: what could change the war by 2035 · Treating the soil rather than the seed (speculative)
• Virtual cell roadmap: from bulk omics to a predictive model of a cancer cell · Atlases and bulk omics (historic)
• Virtual cell roadmap: from bulk omics to a predictive model of a cancer cell · Pooled perturbation screens meet single cells (historic)
• Virtual cell roadmap: from bulk omics to a predictive model of a cancer cell · First single-cell foundation models (current)
• Virtual cell roadmap: from bulk omics to a predictive model of a cancer cell · Data at scale and context-aware models (current)
• Virtual cell roadmap: from bulk omics to a predictive model of a cancer cell · Patient-derived contexts and spatial niches (emerging)
• Virtual cell roadmap: from bulk omics to a predictive model of a cancer cell · Speculative: in silico trials and digital twins (speculative)

• ALASCCA (phase 3, n=626): Time to recurrence at 3 years, group A (PIK3CA exon 9/20): 7.7% vs 14.1%, HR 0.49
• ACE-Breast-02 (phase 3, n=441): Progression-free survival: 11.3 months vs 8.2 months, HR 0.64
• Low-dose olanzapine for cancer anorexia (Tata Memorial) (phase 3, n=124): Weight gain greater than 5% at 12 weeks: 60% vs 9%
• Add-Aspirin (phase 3, n=11,000): Ongoing; primary analyses pending.
• Study to Investigate Outcome of Individualized Treatment in Patients With Metastatic Colorectal Cancer (phase 2, n=40): result not yet recorded
• Atavistic chemotherapy: combinations of antibacterial, antifungal and antiprotozoal drugs in advanced cancer (Arguello clinic, status unknown) (phase 2, n=250): No results posted; status unknown since April 2022 (ClinicalTrials.gov NCT02366884).

1. What is drug-to-antibody ratio and why does it matter?
   Answer

   The number of payload molecules per antibody, typically 2-8. Higher DAR delivers more drug per binding event (T-DXd about 8) but increases hydrophobicity and clearance unless hydrophilic linkers are used.
2. What is DepMap for?
   Answer

   The Broad Institute's Cancer Dependency Map: genome-wide CRISPR and drug screens across more than 1,000 cancer cell lines showing which genes each line cannot live without, the source of synthetic-lethal targets like PRMT5/MTAP.

• NCI: Cancer treatment research: https://www.cancer.gov/research/areas/treatment
• NCI Developmental Therapeutics Program: https://dtp.cancer.gov/