Skip to main content
Home » Antimicrobial Resistance » How generative AI is transforming antibiotic discovery
Sponsored

Akhila Kosaraju, MD

CEO/President, Phare Bio

Scientists are employing the latest advances in artificial intelligence — along with innovative funding approaches — to accelerate the pace of antibiotic discovery.


In drug discovery, AI is becoming a pioneering force for good. For example, Phare Bio, a social venture founded in 2020, uses generative AI to develop novel antibiotics in partnership with Jim Collins’ lab at MIT.

Generative AI is rapidly accelerating antibiotic discovery

“Antibiotics are the backbone of modern medicine, but escalating antimicrobial resistance threatens these critical tools unless novel antibiotics are rapidly discovered and brought to market,” says Dr Akhila Kosaraju, Phare Bio CEO. “By using generative AI to design antibiotics against pathogens on the World Health Organization’s high-priority list, we’re bringing unprecedented speed and novelty to antibiotic discovery and rebuilding the global pipeline.”  

Many pharma companies use predictive AI to streamline drug discovery. This allows them to quickly screen across millions of existing compounds to identify those with properties of interest. Generative AI marks a sea change, however, in designing new compound structures not currently found in nature, enabling more novelty and a new level of disease-targeting.

“We’re currently adding toxicology and formulation filters into our AI platform. With this customisation, we can design antibiotics with reduced toxicities and target product profiles aligned to real patient needs. This would be a gold standard not just for antibiotics but for AI-based drug discovery more broadly,” says Dr Kosaraju.  

Antibiotics are the backbone
of modern medicine.

Hybrid funding model that incentivises drug discovery

It’s not just the discovery process that benefits from innovation. To overcome challenges plaguing antibiotic development, funding approaches need rethinking too. Phare Bio leverages a unique business model to advance candidates through the ‘valley of death,’ the early stage in development when most drugs fail.

“Developing a drug is high-risk and expensive, costing on average $1.3 billion,” says Dr Kosaraju. “That’s why pharma companies tend to focus on products with potential outsized returns, whereas antibiotics, on average, have low profit margins and limited sales. To address this, we employ a hybrid business model using philanthropic and grant funding for the riskiest early-stage preclinical studies, plus commercial partnerships and spinoffs for costly later-stage R&D. In doing so, we aim to rebalance the risk-reward ratio more favourably for antibiotic development.”

Next article