The Drug That Should Have Taken a Decade

The Antibiotic Crisis

Antibiotic resistance kills more than 1.2 million people globally each year. Yet pharmaceutical companies have largely abandoned antibiotic development—the economics don't work when drug development takes 10-15 years and $2.6 billion on average.

MMLI's Breakthrough Approach

The NSF AI Institute for Molecular Discovery and Synthesis (MMLI) is using AI to fundamentally accelerate how we discover new molecules. Their systems can predict molecular properties, identify promising candidates, and suggest synthesis pathways in days rather than years.

For one novel antibiotic candidate targeting resistant bacteria, the AI system screened billions of potential molecular structures, identified thousands of promising candidates, and predicted which would be effective, safe, and synthesizable—work that would have taken human researchers years.

From Lab to Potential Treatment

The candidate compound is now in preclinical development, having reached this stage in a fraction of the typical timeline. If successful, it could help address one of medicine's most urgent challenges.

Key Impacts

90% reduction in initial drug candidate screening time.

Novel compounds identified that human researchers had not considered.