AI cracks superbug mystery in 2 days

Could AI be our secret weapon against superbugs? The shocking results of a two-day breakthrough could change everything.

A recent AI breakthrough, which took just two days to crack what a scientist had been working on for ten years, is having a significant impact.

“Co-scientist,” a tool developed by Google, was used by Professor José R. Penadés to get to the root cause of a longstanding problem: why some superbugs are immune to antibiotics.

In fact, the AI tool went even further, providing four different hypotheses, all of which made sense to the scientists, reported the BBC.

Globally speaking, antibiotic resistance is a looming crisis and traditional research methods often take too long to yield results.

AI could, in fact, be the perfect weapon in the fight against these kinds of superbugs, but it does come with challenges, such as lab validation, unknown biases, and the reluctance of big pharma due to the low profitability of developing new antibiotics.

At the start of the decade, the WHO categorized antimicrobial resistance (AMR) as one of the top 10 global health threats. Without urgent action, drug-resistant infections could cause at least 10 million deaths per year.

Infections like MRSA and TB are becoming increasingly resistant to antibiotics, meaning new research and development are imperative.

When AI enters the battlefield and generates multiple hypotheses, with the top hypothesis confirming the lead scientist’s decade-long idea, it becomes clear that AI is a game-changer in tackling medical challenges.

AI is reshaping medicine by cracking scientific puzzles at record speed. DeepMind’s AlphaFold solved a 50-year-old mystery about protein structures, helping researchers design new drugs.

Insilico Medicine’s AI identified a potential treatment in weeks rather than years, and MIT’s (Massachusetts Institute of Technology) Halicin found an entirely new antibiotic to fight resistant bacteria. These breakthroughs show how AI is accelerating discoveries that could save lives.

AI vs bacteria may not sound like the most orthodox of arms races, but that’s the reality here. In the case of MIT’s Halicin, AI screened thousands of compounds in just days and identified a promising new antibiotic to help combat resistant bacteria.

A huge part of this success lies in predicting how bacteria will evolve, a highly unpredictable process, and develop resistance to antibiotics. By staying one step ahead, scientists might even be able to preemptively develop medicine.

However, there’s a risk that AI’s proposed solutions might not always be safe. AI can’t fully replicate the highly sophisticated tapestry of human biology or accurately diagnose the side effects an individual might experience, as everyone reacts differently.

Even when AI develops a particular antibiotic candidate that sounds good on paper, lengthy testing stages are still necessary to ensure toxicity is kept to a minimum.

Another obstacle is that AI can also be used for more nefarious purposes. A major concern in biosecurity is that technology could be used to design harmful pathogens, including superbugs that are resistant to all known treatments.

So, where could this lead in the future? Well, AI will be able to produce remedies for superbugs much faster than the years it takes scientists to hypothesize.

AI may soon be able to offer personalized treatments for individuals by analyzing their genetic makeup. This means that instead of waiting for time-consuming, broader solutions, the technology could narrow in on specific treatments that could save lives.

When lobbyists call for strict regulation, one of the main areas where AI could really take off is biotech. While thorough governance is necessary, the use of AI in medicine should be heavily encouraged, as shown in the case of Professor Penadés.