Breaking the Wall of Antibiotic Discovery with AI

Breaking the Wall of Antibiotic Discovery with AI  

Global Call 2025 Finalist Interview: Engineering & Technology   

Cesar de la Fuente has pioneered artificial intelligence approaches that have greatly accelerated antibiotic discovery, yielding numerous preclinical candidates that show promise for therapeutic intervention against currently untreatable infections. He has received numerous major awards, including the Princess of Girona Prize, and has published over 180 papers, including in Science and Cell. De la Fuente is a Sloan Fellow and an elected Fellow of the American Institute for Medical and Biological Engineering (AIMBE), becoming one of the youngest ever to be inducted. 

Which wall does your research or project break? 

Breaking the Wall of Antibiotic Discovery with AI: For over a century, finding new antibiotics was a slow, labour-intensive and often serendipitous process. Scientists relied on “dirt mining”–collecting soil microbes and painstakingly screening them for antibiotics–which yielded many early antibiotics but became far less fruitful after the 1970s. The same known compounds were rediscovered repeatedly, and drug pipelines stagnated even as drug-resistant “superbugs” surged. Key barriers included reliance on the tiny fraction of microbes that could be cultured, years-long trial-and-error workflows, costly wet-lab screening and endless dereplication (repeatedly finding molecules that turned out to be already known). This traditional pipeline simply could not keep pace with evolving bacteria. 

My work breaks this wall by replacing serendipity with digital discovery. We integrate artificial intelligence (AI) with biology to turn the entire Tree of Life into a searchable library of potential antibiotics. Instead of hunting in soil, we can scan genomes and proteomes across diverse life forms–from humans and ordinary bacteria to extremophile archaea and even long-extinct creatures–predicting which protein fragments might fight infections. As a result, discoveries that once took years can now happen in hours. 

Notably, my lab pioneered the first computer-designed antibiotic effective in animals, showing that machines can invent new drugs. We also performed the first proteome-wide search of the human body, uncovering over 2,600 “encrypted” antimicrobial compounds hidden within human proteins. Mining the human microbiome, we found novel molecules, including one from gut bacteria that proved as potent as a last-resort antibiotic in preclinical tests. We even tapped into “microbial dark matter,” revealing nearly one million new antibiotic candidates by sifting through tens of thousands of microbial genomes from soil, water and beyond, and we have openly shared these discoveries via our AMPSphere database. 

We also broke the wall of time by reaching into the deep past. In a field we call molecular de-extinction, we digitally resurrected antibiotic molecules from ancient genetic data. For example, by mining Neanderthal and woolly mammoth genomes, we found compounds that no longer exist in nature, synthesized them, and showed they can cure bacterial infections in mice. It is like a real-life “Jurassic Park” for medicine – reaching into the past to retrieve new cures. We even explored the long-ignored domain of archaea and uncovered an entirely new class of antibiotics called archaeasins. In short, we are shattering the old barriers of antibiotic discovery and vastly expanding the arsenal to outpace superbugs. 

What is the main goal of your research or project? 

Building a Digital Antibiotic Discovery Engine: My overarching goal is to create an AI-first, open and scalable engine that compresses antibiotic discovery from years to hours. Ultimately, I aim to generalize this approach across the entire biological world. In practical terms, I want to turn the search for new antibiotics into a predictable, repeatable engineering process instead of a one-off gamble. This means developing algorithms that learn how genetic sequences relate to biological function so they can predict a molecule’s antimicrobial potency, spectrum and safety–or even design new drug candidates from scratch. These AI models can rapidly screen vast chemical and genetic spaces, then output ranked candidate lists for a fast design–build–test cycle in the lab. 

In pursuit of this goal, my team has already achieved key breakthroughs that validate our approach. We designed a new antibiotic with AI that successfully treated bacterial infections in animals, proving that computer-guided drug development works. We then extended our search to unconventional sources: scanning the entire human proteome revealed thousands of hidden antimicrobial fragments in our own bodies, and probing diverse microbiomes uncovered potent molecules from gut bacteria that rival our best antibiotics. We even pushed into unexpected realms–resurrecting extinct compounds (for example, a Neanderthal molecule with therapeutic effects) and mining the archaeal domain to discover a new class of antibiotics called archaeasins. 

Underpinning all these advances is our unified AI platform (APEX) which integrates diverse datasets and continuously learns from each design-test cycle. By automating much of the discovery process, APEX becomes a self-improving engine that gets smarter with every iteration. 

The aim is not just to find a single drug but to establish a durable platform that any researcher can use to rapidly identify and optimize antibiotics on demand. In practice, success would mean a steady pipeline of antibiotic candidates moving into preclinical testing, fuelled by open data, tools and broad collaboration to advance the best leads into clinical trials. Moreover, the same AI framework can be generalized to other therapies–from novel immunomodulators to cancer treatments. By sharing our tools and datasets openly, I want to empower scientists everywhere to respond swiftly to emerging health threats and keep our medicine chest stocked. Ultimately, my goal is to revolutionize how we discover drugs, making the process faster, smarter and more accessible so we can stay ahead of evolving diseases and save lives. 

What advice would you give to young scientists or students interested in pursuing a career in research, or to your younger self starting in science? 

We are all born scientists at heart. I was that kid exploring tide pools, endlessly curious about every creature I found. The key is never to lose that curiosity—guard it fiercely. That sense of wonder will carry you through tough times and fuel your journey. The questions that make others raise their eyebrows are often the ones worth chasing; discovery rarely begins with consensus. 

Don’t be afraid to step into the unknown. I built my career by connecting dots between disciplines, venturing into unfamiliar territory where biology meets computing. Breakthroughs often arise at these crossroads, when diverse ideas collide beyond your comfort zone. If a project feels too comfortable, it is probably not ambitious enough—push the boundaries until today’s limits begin to crack. 

Research is a rollercoaster: experiments fail, ideas hit dead ends. But there is no such thing as failure in science—only data. Each setback is feedback to guide your next step. Persist, iterate and let every stumble sharpen you. Resilience and curiosity together form an unstoppable force. 

Learn to communicate your science clearly and passionately. Whether you are speaking to a child or a Nobel laureate, a vivid explanation multiplies your impact and can spark new ideas in your own mind.  

Science is a team sport, so surround yourself with mentors and colleagues who challenge and inspire you and seek out diverse perspectives. No one solves big problems alone. As you grow, pay it forward: mentor others, bring underrepresented voices to the table and lift up those around you. We all go farther when we go together. 

Defend your focus and your joy. Carve out time for deep work but also know when to step away and recharge. Often a eureka moment strikes when you are away from the lab—hiking or cooking with friends. Balance intense effort with real rest. Keep science fun. 

Keep sight of the bigger purpose. Science isn’t about collecting accolades; it is about expanding knowledge and improving lives. Be audacious and stay curious—keep asking “what if?” until the walls around our hardest questions crumble. If you remain courageous, collaborative and resilient, you won’t just build a rewarding career—you will help build a better world. 

What inspired you to be in the profession you are today? 

I have always been driven by a deep curiosity about how life works and a desire to solve real-world problems. As a child, I would bring home fish and insects to examine—an enduring fascination with nature that, combined with my love of technology and the idea of programming biology like a machine, drew me into research and continues to motivate me every day. 

What impact does your research or project have on society?  

My work aims to accelerate the discovery of new antibiotics using AI, to help combat the global threat of antimicrobial resistance (a crisis projected to cause 10 million deaths per year by 2050 if unchecked). In the long run, I hope this approach also sets a new paradigm for drug discovery that benefits humanity well beyond antibiotics. 

What is one surprising fact about your research or project that people might not know? 

One surprising fact is that we have discovered antibiotic candidates by mining the proteomes of extinct species like Neanderthals and woolly mammoths. In a sense, it is “Jurassic Park” for medicines – a process we call molecular de-extinction–and it opens up new sources for drug discovery that few people realise even exist. 

What’s the most exciting moment you've experienced over the course of your research or project? 

One of the most exciting moments was when our AI predicted an entirely new antibiotic molecule—something never seen before—and then we watched that compound, digitally resurrected from an extinct species, actually cure a bacterial infection in a mouse. It felt like reaching through time to find a cure, proving that our unconventional “time-traveling” strategy was not science fiction but a viable path forward—a validation that even turned sceptics into believers.