Using AI to Find New Antibiotics Still a Work in Progress
Posted on by Lawrence Tabak, D.D.S., Ph.D.
Each year, more than 2.8 million people in the United States develop bacterial infections that don’t respond to treatment and sometimes turn life-threatening . Their infections are antibiotic-resistant, meaning the bacteria have changed in ways that allow them to withstand our current widely used arsenal of antibiotics. It’s a serious and growing health-care problem here and around the world. To fight back, doctors desperately need new antibiotics, including novel classes of drugs that bacteria haven’t seen and developed ways to resist.
Developing new antibiotics, however, involves much time, research, and expense. It’s also fraught with false leads. That’s why some researchers have turned to harnessing the predictive power of artificial intelligence (AI) in hopes of selecting the most promising leads faster and with greater precision.
It’s a potentially paradigm-shifting development in drug discovery, and a recent NIH-funded study, published in the journal Molecular Systems Biology, demonstrates AI’s potential to streamline the process of selecting future antibiotics . The results are also a bit sobering. They highlight the current limitations of one promising AI approach, showing that further refinement will still be needed to maximize its predictive capabilities.
These findings come from the lab of James Collins, Massachusetts Institute of Technology (MIT), Cambridge, and his recently launched Antibiotics-AI Project. His audacious goal is to develop seven new classes of antibiotics to treat seven of the world’s deadliest bacterial pathogens in just seven years. What makes this project so bold is that only two new classes of antibiotics have reached the market in the last 50 years!
In the latest study, Collins and his team looked to an AI program called AlphaFold2 . The name might ring a bell. AlphaFold’s AI-powered ability to predict protein structures was a finalist in Science Magazine’s 2020 Breakthrough of the Year. In fact, AlphaFold has been used already to predict the structures of more than 200 million proteins, or almost every known protein on the planet .
AlphaFold employs a deep learning approach that can predict most protein structures from their amino acid sequences about as well as more costly and time-consuming protein-mapping techniques.
In the deep learning models used to predict protein structure, computers are “trained” on existing data. As computers “learn” to understand complex relationships within the training material, they develop a model that can then be applied for making predictions of 3D protein structures from linear amino acid sequences without relying on new experiments in the lab.
Collins and his team hoped to combine AlphaFold with computer simulations commonly used in drug discovery as a way to predict interactions between essential bacterial proteins and antibacterial compounds. If it worked, researchers could then conduct virtual rapid screens of millions of new synthetic drug compounds targeting key bacterial proteins that existing antibiotics don’t. It would also enable the rapid development of antibiotics that work in novel ways, exactly what doctors need to treat antibiotic-resistant infections.
To test the strategy, Collins and his team focused first on the predicted structures of 296 essential proteins from the Escherichia coli bacterium as well as 218 antibacterial compounds. Their computer simulations then predicted how strongly any two molecules (essential protein and antibacterial) would bind together based on their shapes and physical properties.
It turned out that screening many antibacterial compounds against many potential targets in E. coli led to inaccurate predictions. For example, when comparing their computational predictions with actual interactions for 12 essential proteins measured in the lab, they found that their simulated model had about a 50:50 chance of being right. In other words, it couldn’t identify true interactions between drugs and proteins any better than random guessing.
They suspect one reason for their model’s poor performance is that the protein structures used to train the computer are fixed, not flexible and shifting physical configurations as happens in real life. To improve their success rate, they ran their predictions through additional machine-learning models that had been trained on data to help them “learn” how proteins and other molecules reconfigure themselves and interact. While this souped-up model got somewhat better results, the researchers report that they still aren’t good enough to identify promising new drugs and their protein targets.
What now? In future studies, the Collins lab will continue to incorporate and train the computers on even more biochemical and biophysical data to help with the predictive process. That’s why this study should be interpreted as an interim progress report on an area of science that will only get better with time.
But it’s also a sobering reminder that the quest to find new classes of antibiotics won’t be easy—even when aided by powerful AI approaches. We certainly aren’t there yet, but I’m confident that we will get there to give doctors new therapeutic weapons and turn back the rise in antibiotic-resistant infections.
 2019 Antibiotic resistance threats report. Centers for Disease Control and Prevention.
 Benchmarking AlphaFold-enabled molecular docking predictions for antibiotic discovery. Wong F, Krishnan A, Zheng EJ, Stark H, Manson AL, Earl AM, Jaakkola T, Collins JJ. Molecular Systems Biology. 2022 Sept 6. 18: e11081.
 Highly accurate protein structure prediction with AlphaFold. Jumper J, Evans R, Pritzel A, Kavukcuoglu K, Kohli P, Hassabis D., et al. Nature. 2021 Aug;596(7873):583-589.
 ‘The entire protein universe’: AI predicts shape of nearly every known protein. Callaway E. Nature. 2022 Aug;608(7921):15-16.
Antimicrobial (Drug) Resistance (National Institute of Allergy and Infectious Diseases/NIH)
Collins Lab (Massachusetts Institute of Technology, Cambridge)
The Antibiotics-AI Project, The Audacious Project (TED)
AlphaFold (Deep Mind, London, United Kingdom)
NIH Support: National Institute of Allergy and Infectious Diseases; National Institute of General Medical Sciences
Lyme Disease: Gene Signatures May Catch the Infection Sooner
Posted on by Dr. Francis Collins
Each year, thousands of Americans are bitten by deer ticks.These tiny ticks, common in and around wooded areas in some parts of the United States, can transmit a bacterium into the bloodstream that causes Lyme disease. Those infected experience fever, headache, stiff necks, body aches, and fatigue. A characteristic circular “target” red rash can mark the site of the tick bite, but isn’t always noticed. In fact, many people don’t realize that they’ve been bitten, and weeks can pass before they see a doctor. By then the infection has spread, sometimes causing additional rashes and/or neurological, cardiac, and rheumatological symptoms that mimic those of other conditions. All of this can make getting the right diagnosis frustrating, especially in areas where Lyme disease is rare.
Even when Lyme disease is suspected early on, the bacterium is unusually slow growing and present at low levels, so it can take a while before blood tests detect antibodies to confirm the condition. By then, knocking out the infection with antibiotics can be more challenging. But research progress continues to be made toward improving the diagnosis of Lyme disease.
An NIH-supported team recently uncovered a unique gene expression pattern in white blood cells from people infected with the Lyme disease-causing bacterium Borrelia burgdorferi . This distinctive early gene signature, which persists after antibiotic treatment, is unique from other viral and bacterial illnesses studies by the team. With further work and validation, the test could one day possibly provide a valuable new tool to help doctors diagnose Lyme disease earlier and help more people get the timely treatment that they need.
Gene Expression Test Aims to Reduce Antibiotic Overuse
Posted on by Dr. Francis Collins
Without doubt, antibiotic drugs have saved hundreds of millions of lives from bacterial infections that would have otherwise been fatal. But their inappropriate use has led to the rise of antibiotic-resistant superbugs, which now infect at least 2 million Americans every year and are responsible for thousands of deaths . I’ve just come from the World Economic Forum in Davos, Switzerland, where concerns about antibiotic resistance and overuse was a topic of conversation. In fact, some of the world’s biggest pharmaceutical companies issued a joint declaration at the forum, calling on governments and industry to work together to combat this growing public health threat .
Many people who go to the doctor suffering from respiratory symptoms expect to be given a prescription for antibiotics. Not only do such antibiotics often fail to help, they serve to fuel the development of antibiotic-resistant superbugs . That’s because antibiotics are only useful in treating respiratory illnesses caused by bacteria, and have no impact on those caused by viruses (which are frequent in the wintertime). So, I’m pleased to report that a research team, partially supported by NIH, recently made progress toward a simple blood test that analyzes patterns of gene expression to determine if a patient’s respiratory symptoms likely stem from a bacterial infection, viral infection, or no infection at all.
In contrast to standard tests that look for signs of a specific infectious agent—respiratory syncytial virus (RSV) or the influenza virus, for instance—the new strategy casts a wide net that takes into account changes in the patterns of gene expression in the bloodstream, which differ depending on whether a person is fighting off a bacterial or a viral infection. As reported in Science Translational Medicine , Geoffrey Ginsburg, Christopher Woods, and Ephraim Tsalik of Duke University’s Center for Applied Genomics and Precision Medicine, Durham, NC, and their colleagues collected blood samples from 273 people who came to the emergency room (ER) with signs of acute respiratory illness. Standard diagnostic tests showed that 70 patients arrived in the ER with bacterial infections and 115 were battling viruses. Another 88 patients had no signs of infection, with symptoms traced instead to other health conditions.