![\"Coloured](\"https:\/\/www.newsbeep.com\/uk\/wp-content\/uploads\/2026\/06\/d41586-026-01818-9_52511298.jpg\"\/)
<\/p>\n <\/p>\n
Some bacteria are becoming resistant to the antibiotics commonly used to control them.Credit: NIAID\/NIH\/Science Photo Library<\/p>\n
When it comes to bacterial infections in the gut, antibiotics are effective yet broad-acting: although they might kill disease-causing species effectively, beneficial microflora can get caught in the crossfire. This indiscriminate effect can be harmful, particularly for people with Crohn\u2019s disease or other chronic gastrointestinal conditions. It also increases the likelihood that antibiotic-resistant strains of bacteria will evolve.<\/p>\n
In 2023, microbiologist Jonathan Stokes at McMaster University in Hamilton, Canada, began looking for options that could target pathogens with greater precision. He and his colleagues screened some 10,000 bioactive compounds for antibacterial activity against a strain of Escherichia coli that can cause severe gut infections. They filtered the results on the basis of various criteria, including toxicity to bacteria and structural novelty compared with existing antibiotics. \u201cWe got super lucky in that we only ended up with one molecule,\u201d says Denise Catacutan, the doctoral student in Stokes\u2019s laboratory who led the work1<\/a>.<\/p>\n But the team needed to confirm that the promising molecule, named enterololin, was specific to its target pathogen, rather than acting as another broad-spectrum antibiotic. Typically, researchers rely on extensive biochemical screens, RNA sequencing or proteomics to elucidate how molecules such as enterololin disrupt bacterial pathways. This time, the team turned to artificial intelligence.<\/p>\n How we\u2019re using AI tools to improve psychedelic-drug research<\/p>\n <\/a><\/p>\n AI tools can accelerate the process of developing pharmaceuticals and are fast becoming a crucial component of drug discovery. But using AI to identify an antibiotic\u2019s mechanism of action is still uncommon, says Regina Barzilay, a computer scientist at the Massachusetts Institute of Technology (MIT) in Cambridge.<\/p>\n Barzilay\u2019s lab developed a tool to fill that gap. DiffDock<\/a> uses AI to predict how small molecules bind to proteins. It can thereby identify possible protein targets \u2014 and potential mechanisms of action for the small molecules. By applying this tool to enterololin, Stokes and his research group \u201ccould kind of narrow down our experimental pipeline\u201d, Catacutan says. The team developed bacterial strains with mutations in the genes encoding predicted target proteins, and quickly confirmed DiffDock\u2019s predictions.<\/p>\n Barzilay\u2019s interest in antibiotics is personal. Her father contracted a bacterial infection of the spine that required complex surgery, and another family member survived an infection that didn\u2019t respond to any antibiotics. \u201cWe are so used to the idea that antibiotics are there to protect us,\u201d she says. But that protection is precarious. Antibiotic resistance is a pervasive, growing global crisis; estimates suggest that drug-resistant infections could kill at least 39 million people by 2050.<\/p>\n Yet antibiotic development and manufacturing are expensive and rarely profitable, so drug companies are reluctant to invest. Discovering antimicrobials that can be synthesized easily \u2014 and cheaply \u2014 could help. An increasing number of researchers are turning to AI to address this need. Using machine-learning tools to tackle tasks in silico, from identifying new antibiotic candidates to predicting potential mechanisms of action, enables researchers to work faster \u2014 and on tighter budgets.<\/p>\n Strong foundation<\/p>\n Barzilay\u2019s interest in antibiotic development began in 2018, when she met MIT biomedical engineer James Collins at an institution-wide symposium on the use of AI tools. Barzilay and Collins teamed up to apply these techniques to antibiotic discovery. Stokes, who was a postdoctoral researcher in Collins\u2019s lab with expertise in high-throughput screening of small molecules, joined the effort.<\/p>\n The team developed a model based on neural networks (machine-learning architectures inspired by the human brain) that correlated molecular features \u2014 for example, bond types, atomic number and electronic charge \u2014 with properties such as solubility and microbial growth inhibition.<\/p>\n The researchers trained their model \u2014 called Chemprop<\/a> \u2014 on data from 2,300 or so molecules that had been tested for their ability to inhibit the growth of E. coli. They then used the model to screen millions of molecules for potential drug candidates, eventually homing in on a kinase inhibitor that the group named halicin. This proved to have a potent effect on several pathogenic species, including Mycobacterium tuberculosis (the causative agent of tuberculosis); drug-resistant E. coli; and Acinetobacter baumannii (an opportunistic pathogen that can cause infections in hospitalized individuals)2<\/a>. \u201cWe were able to create a model that can generalize to totally unseen classes of chemistry,\u201d Barzilay says.<\/p>\n But having training data is only a first step \u2014 how it\u2019s labelled and classified is equally important, says Molly Bartlett, a chemical informatician at Imperial College London. Bartlett works with the Fleming Initiative, a multi-institutional collaboration headed by Imperial College London and Imperial College Healthcare NHS Trust that aims to combat antimicrobial resistance globally.<\/p>\n When compiling the initiative\u2019s data, Bartlett searched the published literature for examples of high-throughput screening for molecules that can breach the outer cell membrane of pathogenic bacteria and accumulate inside the cells. Bartlett uses RDKit and xTB, computational tools that simulate and analyse molecular structures, to mimic how these molecules might behave when dissolved in water or in a cell\u2019s lipid membrane. She also categorizes the chemical features that are responsible for various properties of the molecules, such as their solubility and ability to enter the bacterial cell. \u201cIf the input is not able to represent what makes the property happen, you\u2019re not going to be able to get a good prediction,\u201d she explains.<\/p>\n Computer scientist Regina Barzilay is trying to use AI to decipher drugs\u2019 mechanisms.Credit: Sophie Park for The Washington Post via Getty<\/p>\n In Stokes\u2019s experience, a strong training data set should include at least a subset of available clinical drugs, as well as potential antibiotics that are not currently used in the clinic. Training data should also be physically, chemically and structurally diverse \u2014 and should represent powerful antimicrobials, as well as ineffective ones, so that AI models can also learn what not to do. For those interested in building AI tools to find new antibiotics, \u201c80% of your time has to be spent on data acquisition, data processing and data representation,\u201d Stokes advises.<\/p>\n Bartlett, for instance, says that at least 10% of the molecules in her training data need to penetrate the bacterial envelope, so that the model can learn the characteristics that predict accumulation. Some databases she works with contain upwards of 100,000 molecules, but often only 3% of those molecules can enter the type of bacteria that interests her. She also tries to maximize the diversity of chemical structures represented in the training data. Without this breadth, she says, \u201cyou\u2019re not going to be able to have a predictive model\u201d.<\/p>\n Generative AI tools have made her work easier. Bartlett and Catacutan struggled at first with writing the code necessary to run models, but now use AI tools such as Google\u2019s Gemini and OpenAI\u2019s ChatGPT to help with troubleshooting. Bartlett will occasionally provide Gemini with a toolkit\u2019s explanatory text file, often called a README file, then give the chatbot detailed instructions on what errors to look for in her code. \u201cYou still have to know what to ask for, but you don\u2019t have to be an expert in the architecture itself of the code,\u201d she says. \u201cIt makes it really accessible.\u201d<\/p>\n Peptide power-up<\/p>\n Also on a quest for chemical diversity is C\u00e9sar de la Fuente, a synthetic biologist at the University of Pennsylvania in Philadelphia. De la Fuente\u2019s work focuses on antimicrobial peptides: naturally occurring short chains of amino acids that could prove to be potent antibiotics. His lab has also pioneered a technique called molecular de-extinction, which aims to \u2018resurrect\u2019 molecules that might have useful biological characteristics from extinct organisms.<\/p>\n Six key developments in the fight against antimicrobial resistance<\/p>\n <\/a><\/p>\n De la Fuente and his team used a combination of neural networks to create a tool called APEX<\/a> (antibiotic peptide de-extinction). They used this to screen a database of more than ten million peptides, and identified more than 37,000 that were predicted to have broad-spectrum antimicrobial activity3<\/a>. About 11,000 of these were derived from the \u2018extinctome\u2019 \u2014 proteomes from extinct creatures, including, in this case, an ancient magnolia, a giant sloth and a Grant\u2019s zebra.<\/p>\n The researchers synthesized and tested 69 of these candidate molecules against bacterial pathogens and found that many of them had an unusual mechanism of action. Rather than targeting a pathogen\u2019s outer, rigid cell wall, these compounds acted on the inner cytoplasmic membrane \u2014 a strategy that might make them more robust as antibiotics, because bacteria are less likely to have evolved resistance to molecules they have never encountered. \u201cEvolution is this beautiful planetary-scale optimization process,\u201d de la Fuente says. It\u2019s \u201cthe biggest optimization experiment that we have ever seen\u201d.<\/p>\n Making molecules<\/p>\n Having learnt from extinct molecules, de la Fuente\u2019s team took the next step: they developed a generative AI model that can design synthetic molecules that don\u2019t \u2014 yet \u2014 exist in nature4<\/a>. \u201cGenerative AI gives you this opportunity now to go beyond the sequence space that evolution has explored, to come up with things that may have properties and functions that are more optimized in certain ways,\u201d de la Fuente says.<\/p>\n The team provides its model \u2014 called ApexGO \u2014 with a peptide template, and specifies design goals, constraints and rules, such as how closely the designs must adhere to the starting peptide. People in the lab then step in to assess which designs could be viable. They might, for instance, exclude peptides that have too many hydrophobic residues, which would cause clumping in solution, de la Fuente says. The researchers then synthesize promising molecules, which they study in cultured cells and in animal models of infection. The team has so far synthesized and tested around 100 peptides, de la Fuente says. About 86% showed antimicrobial activity against at least one pathogen.<\/p>\n![\"\"](\"https:\/\/www.newsbeep.com\/uk\/wp-content\/uploads\/2026\/06\/d41586-026-01818-9_52516170.jpg\"\/)
<\/p>\n![\"Portrait](\"https:\/\/www.newsbeep.com\/uk\/wp-content\/uploads\/2026\/06\/d41586-026-01818-9_52511296.jpg\"\/)
<\/p>\n![\"\"](\"https:\/\/www.newsbeep.com\/uk\/wp-content\/uploads\/2026\/06\/d41586-026-01818-9_52379200.jpg\"\/)
<\/p>\n
![\"\"](\"https:\/\/www.newsbeep.com\/uk\/wp-content\/uploads\/2026\/06\/d41586-026-01818-9_52398734.jpg\"\/)
<\/p>\n