A nonprofit working to keep the medicines that fight fungal infections working.
For a world without antifungal resistance.
Fungal infections kill an estimated 3.8 million people a year. The handful of medicines we have to fight them are slowly stopping working, because the fungi are learning to survive them. Almost no one is working on this. We build free, carefully checked tools to help keep those medicines useful, and to track the dangerous infections that are starting to resist them.
What we're building
A free, live tracker for a drug-resistant hospital fungus.
Candida auris is a fungus that spreads in hospitals, and it is increasingly surviving the drugs meant to kill it. Labs around the world share its DNA in public databases. We built a free tool that reads that DNA, spots the exact genetic changes that let the fungus resist one important group of drugs, and shows the current picture on a public page anyone can check.
A tracker is only as trustworthy as the tool behind it. Before we trusted ours, we tested it on 98 cases that labs had already worked out by hand. It got all 98 right: it never missed a resistant case and never raised a false alarm. An automatic check re-runs that same test every time we change the code, so it stays honest.
The live page shows how many cases we're seeing right now and how many are resistant. Those numbers change as new DNA is shared, so we don't freeze a count here.
Why this matters
These infections are rising, harder to treat, and mostly ignored.
A hidden killer
Serious fungal infections kill millions of people a year, usually people already weakened by cancer, HIV, or a stay in intensive care. They get a tiny fraction of the attention that bacterial infections get.
The drugs are starting to fail
A fungus called Candida auris is spreading through hospitals and shrugging off several of our antifungal drugs at once. The World Health Organization lists it as one of the most urgent fungal threats.
Almost nothing new is coming
Very little money goes into new antifungal drugs. Few new ones are on the way, and the ones we already have are slowly losing their punch.
What we do
Learn how fungi work. Build in the open. Prove it before we claim it.
Study the biology
We dig into how fungi cause disease and how they survive the drugs meant to kill them.
Build free, open tools
Everything we make is free and open, so any lab or hospital can use it and check it, not just us.
Connect the people who can act
We link up the computer work with the labs, doctors, and funders who can turn an idea into a real treatment.
How the drugs work
It really comes down to shape.
Scroll to see how an antifungal drug actually kills a fungus, and what makes the difference between a molecule that works and one that only looks like it should.
Follow the molecule
Meet the target
Deep inside a fungal cell sits a tiny machine the fungus needs to survive. At its center is a single iron atom.
How we work
We prove our own tools before we ask anyone to trust them.
A computer prediction is just a guess until someone checks it, and it's easy to show off a confident-looking result that means nothing. So we test ours against answers we already know are right before we let them loose on anything new. We publish everything, including the tries that fail.
Also in the works
Can a computer spot a promising new drug before anyone tests it in a lab?
Alongside the tracker, we're testing a bolder idea: using the physical shape of a molecule to judge whether it could become a new antifungal drug, instead of relying on the usual scoring programs. As a fair test, we hid 7 real antifungal drugs among 348 look-alike molecules and asked our method to pick them out, with the rules locked in beforehand so we couldn't nudge the result.
It worked. Real drugs rose to the top far more often with our method than with the standard program, on the exact same data, and the odds this was just luck were about 3 in 1,000. It's early, and it's only one test, but it's real evidence the idea has legs.
What this does not prove yet
- We only had 7 real drugs to find, so we can't be fully sure yet.
- The look-alike molecules are assumed to be duds, not confirmed in a lab.
- We used a simplified, fixed model of the target, not a fully lifelike one.
- This points to ideas worth testing. It doesn't find a drug. A high score is a lead for a lab, never a cure.
Where we're headed
From a tool we've proven to open tracking anyone can use.
Prove the tool works
DoneWe checked our resistance detector against known lab results and set up an automatic test that re-checks it every time we change the code.
Put it online for everyone
DoneWe turned it into a free, always-updating public page that shows where this drug-resistant fungus is turning up.
Cover more, update more often
UnderwayWatch a wider slice of the data, refresh it more often, and use the same careful approach for more infections and more drugs.
Open it up as a shared service
UnderwayGrow into free data and tools any lab or health team can build on, so a warning reaches the people who can act on it.
Test new drug ideas in the lab
Seeking partnersPut our most promising molecules in front of real fungus. We can't do this part alone: it needs lab partners, and nothing is proven until they run it.
Help us keep these medicines working.
Whether you run a lab, study fungi, fund work on neglected diseases, or just want to help with open science, there's a place for you here. Say hello.