Skip to content
Home » From Willow Bark to Algorithms: How Are Drugs Discovered?

From Willow Bark to Algorithms: How Are Drugs Discovered?

    For most of history, the answer to “where do medicines come from?” was nature. It still is, more than most people realize. What has changed is how we search.

    First, folk remedies

    Willow bark was chewed for pain and fever as far back as Hippocrates. In 1828 a German pharmacist isolated its active ingredient, and in 1897 Bayer chemists made a gentler version that the company named Aspirin in 1899.

    Foxglove has a similar story. English physician William Withering learned of an old woman’s herbal tea for swelling, tracked down the active plant, and published a study of 158 patients in 1785. Peruvians used cinchona bark for fevers long before French chemists isolated quinine from it in 1820.

    Sometimes the old knowledge is still paying off. Tu Youyou won the 2015 Nobel Prize after combing ancient Chinese texts and finding that sweet wormwood, used for around 2,000 years, led to artemisinin, a major malaria drug.

    Then, dirt

    In 1943, Rutgers scientist Selman Waksman and his students isolated streptomycin after nearly three decades of studying soil microbes. That kicked off what’s called the golden age of antibiotics. Drug companies asked missionaries, explorers, and pilots to bring back soil from distant places. Pfizer’s Terramycin came from an Indiana soil sample. By the end, companies kept finding the same compounds again, and the dirt looked tapped out.

    Nature is still in the mix

    A review of drug approvals from 1981 to 2019 found that about half of new small-molecule drugs were natural products or built on natural structures.

    The weight-loss drugs are a recent example. In 1992, endocrinologist John Eng found a hormone in Gila monster venom that resembled a human gut hormone but lasted far longer. His employer wouldn’t patent it, so he did, and the first drug from it, exenatide, reached the market in 2005. Today’s blockbusters, such as semaglutide and tirzepatide, are not made from lizard venom. They’re lab-built versions based on human hormones. They are now among the best-selling drugs in the world: in 2025, Lilly’s tirzepatide drugs together brought in about $36.5 billion, more than Merck’s Keytruda.

    Now, algorithms

    The newest search tools are computational. AlphaFold, an AI that predicts protein shapes, helped earn its creators a share of the 2024 Nobel Prize in Chemistry. Insilico Medicine says its drug rentosertib had both its target and its molecule found using AI, and a Phase 3 trial began in 2026.

    A reality check: as of this fall, no AI-discovered drug has been approved, and AI-designed drugs still fail in mid-stage trials about as often as other drugs. AI speeds up the search. It can’t skip the human tests.

    Why it matters to me

    If the search gets faster, the questions in my other posts, like drugs that rewire the brain, may arrive sooner than we expect. They’re also the heart of my novel, Syncope. An early scene has Tom Glass and a friend over coffee debating whether an algorithm can find medicines in the field faster than a century of digging in dirt did.

    Share this post: Reddit · WhatsApp · Email

    Leave a Reply

    Your email address will not be published. Required fields are marked *