Drug-resistant infections kill more than a million people annually, but big pharmaceutical companies have largely abandoned antibiotic research due to high costs. Autumn Arnold, a PhD student in the Stokes Lab at McMaster, developed ESKAPE to address this. The model is free to use: researchers upload a list of compounds and receive predictions of antibacterial activity. "All you need is an internet connection," said microbiologist Jonathan Stokes, who runs the lab. Researchers retain ownership of their discoveries and do not have to share the compounds they submit. So far, more than 12 million molecules have been screened using ESKAPE, which can evaluate over 40,000 compounds a day. Separately, researchers at MIT and Sweden's Karolinska Institutet are using generative deep learning to create entirely new molecules targeting antibiotic resistance. "This is a global problem," Stokes said. "We need everybody helping out."
Free AI tool speeds up search for new antibiotics
ESKAPE model screens thousands of compounds daily to fight drug-resistant infections
The Daily Edition
The front page in your inbox each weekday morning: the lead, the picks, and why each one earned its place.
Analysis
Why This Matters
- Drug-resistant infections are a growing global health crisis, killing over a million people each year.
- The high cost of antibiotic R&D (around $1 billion per drug) has driven big pharma away, leaving the pipeline thin.
- A free, accessible AI tool like ESKAPE could democratise antibiotic discovery and speed up the search for new treatments.
Background
Antibiotic resistance occurs when bacteria evolve to withstand the drugs designed to kill them, making infections harder to treat. The World Health Organization has called it one of the top global public health threats. The development of new antibiotics has slowed dramatically as economic returns are low compared to other drugs, prompting the need for innovative, low-cost discovery methods.
Key Perspectives
Jonathan Stokes, microbiologist: Describes a "perfect storm with rising resistance, a really lean antibiotic pipeline, and enormous economic headwinds", and says the goal is to make antibiotic research accessible to scientists everywhere.
What to Watch
- Adoption of the ESKAPE model by research labs worldwide and any new antibiotics discovered through it.
- Whether the generative deep learning approach from MIT and Karolinska leads to clinical candidates.
- Policy responses to incentivise antibiotic development, such as government subsidies or pull incentives.
Sources
Checked
12 claims in this story checked against 1 source: 9 supported.
Show each claim
- Partly supportedThe free AI tool speeds up the search for new antibiotics.Semafor: “which has the potential to dramatically speed up drug discovery.”
- SupportedESKAPE screens thousands of compounds daily to help fight drug-resistant infections.Semafor: “In total, the model can screen more than 40,000 compounds a day, which has the potential to dramatically speed up drug discovery.”
- Not in the sourcesResearchers at McMaster launched ESKAPE.Semafor: “Autumn Arnold, a PhD student in the Stokes Lab at McMaster University in Canada, came up with a fix: an AI model”
- SupportedESKAPE predicts which chemical compounds might kill dangerous bacteria.Semafor: “an AI model that predicts which chemical compounds might kill dangerous bacteria.”
- SupportedDrug-resistant infections kill more than a million people each year.Semafor: “Drug-resistant infections kill more than a million people each year.”
- Partly supportedBig pharmaceutical companies have largely abandoned antibiotic research because of high costs.Semafor: “But big pharmaceutical companies have largely abandoned the quest for new antibiotics, because the research and development needed to discover them is time-consuming and expensive.”
- SupportedAutumn Arnold, a McMaster PhD student in the Stokes Lab, developed ESKAPE.Semafor: “Autumn Arnold, a PhD student in the Stokes Lab at McMaster University in Canada, came up with a fix: an AI model”
- SupportedJonathan Stokes said, “All you need is an internet connection.”Semafor: ““All you need is an internet connection,” said Jonathan Stokes, a microbiologist who runs the lab.”
- SupportedResearchers retain ownership of their discoveries and do not have to share submitted compounds.Semafor: “Researchers retain ownership of their discoveries, and don’t have to share the compounds they submit.”
- SupportedMore than 12 million molecules have been screened using ESKAPE.Semafor: “So far, researchers have screened more than 12 million molecules using the ESKAPE model.”
- SupportedESKAPE can evaluate over 40,000 compounds a day.Semafor: “In total, the model can screen more than 40,000 compounds a day”
- SupportedStokes said, “This is a global problem. We need everybody helping out.”Semafor: ““This is a global problem,” Stokes says. “We need everybody helping out.””