Drug Discovery Platforms

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.

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 platforms are the GPUs, model libraries, and cloud services that pathology, radiology, and drug-design AI run on.

Using machine learning to pick targets, design molecules and antibodies, and predict which ADC will work.

Predicts the 3D shape of proteins together with DNA, RNA, small molecules and antibodies, the starting point for much modern drug design.

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 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.

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.

Freezers full of consented tumour samples with matched clinical data, which every biomarker and drug programme depends on.

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.

Chemotherapy doses are usually written per square metre of body surface, a convention from 1958 that scales drug clearance between species and people; it is imprecise, and for many newer drugs flat or weight-based doses have replaced it.

Open-source structure models that match AlphaFold 3, with Boltz-2 also predicting how strongly a drug binds.

Large public collections of cancer cell lines profiled for their genomes, drug sensitivity and gene dependencies, the reference data behind much of modern target discovery.

Structure and antibody-design models from Chai Discovery, with Chai-2 reporting high zero-shot antibody hit rates.

Generative chemistry platform behind the first AI-discovered drug to reach phase 2, plus oncology candidates.

Peter Nowell's 1976 idea that a tumour is an evolving population of competing clones is now measured directly by sequencing several regions or repeated blood samples, and models of branching evolution predict which clones will drive relapse.

Knocking out every gene one at a time in cancer cells to find which ones they cannot live without.

Designing a protein from scratch on a computer to grip a chosen target, instead of finding one in an animal or a library.

Using a model of what would have happened to a patient on standard treatment, so fewer people have to be randomised to it.

Folded DNA machines that open only when they touch a tumour, releasing a payload or clotting the tumour's blood supply.

Bacteria that seek out the low-oxygen core of tumours, then manufacture a drug on the spot.

Loading the tiny vesicles cells naturally use to talk to each other with a cancer drug, so the body treats the carrier as its own.

ESM3 is a generative protein model that designed a working fluorescent protein far from any natural sequence.

Mathematics from population genetics shows that resistant cells almost always exist before treatment in large tumours, and that combining drugs with different resistance mutations from the start can succeed where the same drugs in sequence fail.

Cancer cells are treated as players whose success depends on what neighbouring cells do, which lets researchers predict how a tumour's mix of cell types shifts under treatment and design schedules that steer it.

Blood cancer cells taken from a patient's blood or marrow are exposed within days to a panel of approved drugs, and the ones that kill the cancer cells while sparing healthy ones are offered back to the patient; a Vienna trial found this beat the previous treatment in more than half of heavily treated patients.

Dog dewormers and anti-parasite drugs are promoted on social media as hidden cancer cures on the strength of cell-culture experiments and anecdotes. No clinical trial shows benefit in people, liver damage has been reported, and drug repurposing is real but works through trials, not forums.

Growing a patient's own cancer cells in a dish and testing drugs on them directly, instead of guessing from genetics.

Geneformer is a transformer trained on about 30 million single cells that encodes each cell as a ranked list of its genes, so deleting a gene in silico shows which genes matter in a disease. It was the first single-cell foundation model in general use, though benchmarks find only modest gains over linear baselines on some tasks.

Mice carrying the same mutations as human cancers so that tumours arise in the right tissue with an intact immune system; the KPC pancreatic model is the best known.

Cloud systems where hospitals and researchers store and analyse genomic data securely at petabyte scale.

Every cancer drug plant works under Good Manufacturing Practice rules and is inspected. When inspectors find problems they write them up (a Form 483 in the United States), and if the answers are poor a public warning letter, an import ban or a court order can follow. Those actions protect patients, and they are also how shortages begin.

Resistant cells arise by chance mutation as a tumour grows, so the chance of a cancer already containing resistant cells rises with its size. The model argued for treating early and for alternating non-cross-resistant drugs.

Tumours do not grow exponentially forever: growth slows as they enlarge, following a curve Benjamin Gompertz devised for human mortality in 1825 and Anna Kane Laird fitted to tumours in 1964. It explains why small tumours are the most chemosensitive and why doubling times lengthen.

Testing millions or billions of chemical compounds against a cancer target automatically to find starting points for new drugs.

Immunodeficient mice given a human immune system from stem cells, so human immunotherapies and CAR-T cells can be tested against human tumours in a living animal.

In vivo base and prime editing would rewrite a cancer's DNA letter by letter inside the body. It works in the liver for inherited disease; nobody has yet corrected a cancer this way in a person.

A dose of chemotherapy kills a constant fraction of cancer cells, not a constant number, so each cycle removes the same proportion, which is why treatment continues after the tumour has disappeared from scans.

Long-read sequencing (PacBio HiFi, Oxford Nanopore) reads single DNA molecules in stretches of thousands of bases, so rearrangements, repeat expansions, gene fusions and methylation appear in one run where short-read machines miss them. Nanopore can classify a brain tumour during surgery in under an hour; throughput per dollar still trails the largest short-read instruments.

Mathematical oncology writes down how tumours grow, evolve, respond to treatment and interact with the immune system as equations or simulations, then uses them to design doses, schedules and trials. The models themselves are records, each with what it was fitted to and what it was used to decide.

From Paget's seed-and-soil idea to models that estimate when metastases were seeded from a primary and how long they lay dormant, these frameworks explain late relapse and argue for treating micrometastases early.

Molecular glues are small molecules that stick two proteins together so the cell destroys one of them. They are smaller and more drug-like than bifunctional degraders.

Building a trial around one patient, or letting one trial swap drugs in and out as evidence accumulates.

Because tumours regrow fastest when small, the best way to finish them is to give the same chemotherapy doses closer together. The idea, from Larry Norton and Richard Simon, was proved in breast cancer by the CALGB 9741 trial and made two-weekly chemotherapy a standard.

NVIDIA BioNeMo is a software stack, not a model: GPU-optimised training recipes and inference services on which protein, DNA and single-cell foundation models such as Evo 2, ESM and Geneformer are trained and served. The Arc Institute, Recursion and pharmaceutical companies use it, at the price of being tied to NVIDIA hardware and tooling.

Organoid-guided therapy means routinely growing a piece of each patient's tumour and testing drugs on it before choosing, rather than relying on genetics alone.

A digital twin is a computer model of one patient's tumour and body, updated with each scan and blood test, used to forecast how the disease will respond to each option before it is tried.

Patient-derived organoids are miniature 3D versions of a patient's tumour grown in the lab.

A patient-derived xenograft is a patient's tumour grown in a mouse, used to test drugs before they reach people.

Payload-linker synthesis makes the cytotoxic small molecules inside ADCs (exatecan, MMAE, DM1, PBD dimers), whose occupational exposure limits sit in the nanogram range, in facilities built so a speck of dust cannot harm a worker. A handful of licensed sites such as Lonza and WuXi STA supply them, and their lead times gate hundreds of ADCs in development.

Growing a patient's pancreatic tumour as mini-organs in a dish and testing chemotherapies on them to pick the regimen most likely to work.

Equations that describe how a drug's concentration rises and falls in the body and how that concentration translates into effect and toxicity; the reason doses are given per square metre, why some drugs are infused over days, and how children's doses are set.

A model trained on billions of cell microscopy images to read what a drug or gene knockout does to a cell.

Programmable DNA-targeting therapeutics are an experimental idea: a drug that reads a cell's DNA, recognises a cancer-specific sequence, and kills only cells that carry it. Change the guide, and the same drug becomes a new drug.

Gliomas grow by both dividing and migrating through the brain, and a two-parameter equation fitted to a patient's MRI scans estimates how far invisible cells have spread, which can guide how much brain to irradiate and how fast the tumour will grow.

Measuring the proteins in a tumour, which is what drugs actually hit, rather than the genes that encode them.

Machines that measure thousands of proteins at once from tissue or blood, used to find drug targets and early-detection markers.

Mechanistic computer models that join a drug's pharmacology to the biology of the tumour and the body, used by developers and regulators to pick doses, predict combinations and explain why a trial failed.

The speed at which a molecular marker falls during treatment predicts outcome better than a single level: BCR-ABL halving time in chronic myeloid leukaemia and circulating tumour DNA slopes in solid tumours are now used to judge response within weeks.

The tools that design entirely new proteins to bind a chosen target, now used for cancer binders and antibodies.

A GPT-style model for single-cell data that predicts cell types, perturbation responses, and gene networks.

RNA drugs that copy themselves inside the cell, or are made as a loop so they last longer. Both aim to get more protein from a smaller dose.

Reading the genes of each individual cell, and mapping where each cell sits in the tumour.

Site-specific conjugation and linker chemistry decide exactly where and how many payloads attach to the antibody, which determines how safe and effective an ADC is.

Before a tablet or a vial exists, the active drug itself has to be made: many chemical steps in reactors, or extraction from a plant, at a handful of factories most patients never hear of. When one of those factories stops, whole cancer drugs disappear.

Growing tumours compress themselves and their surroundings; models of this solid stress explain collapsed vessels, poor drug delivery and stiff stroma, and point to drugs that soften the tumour so treatment can get in.

Spatial biology instruments are machines that map which genes and proteins are active in each part of a tumour slice.

Methods that read which genes are switched on in each spot or cell of a tumour slice while keeping the tissue's geography, so scientists can see how cancer, immune and stromal cells sit next to one another.

Predicts how cells will respond to a drug or gene knockout, trained on over 100 million perturbed cells.

STRIDE is a microscope test that lights up individual broken DNA strands inside cells, so a laboratory can count how much DNA damage a tumour carries or a drug causes, cell by cell.

Structural biology infrastructure is the microscopes, X-ray sources, and prediction models that show what a cancer protein looks like so chemists can design a drug to fit it.

Finding a second gene that a cancer needs only because its first gene is broken, then hitting the second one.

Testing cheap old drugs, aspirin, metformin, statins, beta-blockers, as cancer treatments, because they are safe, available and sometimes work.

Some tumours contain bacteria and fungi that shelter cancer cells and break down chemotherapy. Killing them may make treatment work.

Stiff, high-pressure tumours squeeze their own blood vessels shut, keeping drugs out. Softening them is a way in.

Radiotherapy works through repair, redistribution, reoxygenation and repopulation between fractions; the discovery that tumours speed up their regrowth during a course explained why gaps in treatment cost cures and led to accelerated schedules.

Sequencing several parts of a tumour, or blood over time, to draw its family tree of mutations and see which branches drive relapse and resistance.

Curves that turn a radiation dose into a probability: how likely the tumour is to be eradicated and how likely a nearby organ is to be damaged. They underlie dose constraints, dose escalation trials and comparisons between treatment plans.

How long a tumour takes to double in volume, measured from two scans; it separates cancers from benign nodules in lung screening, sorts aggressive from indolent disease and estimates how long a tumour has been present.

Predator-prey style equations describe how immune cells hunt tumour cells, and they reproduce dormancy, escape and the delayed, sometimes explosive, responses seen with immunotherapy; they now help design combination and scheduling trials.

Small microfluidic devices that grow tumour cells with blood-vessel-like channels and immune cells, letting researchers watch drugs act in a more life-like setting than a dish.

Producing the engineered viruses that carry a CAR gene into T cells. Viral vector manufacturing is a long-standing bottleneck for cell and gene therapy.

Reading all the genes (exome) or the entire DNA (genome) of a tumour, rather than a chosen panel.