TLDR: Researchers at MIT and the Broad Institute have leveraged generative AI to design two novel antibiotic compounds, NG1 and DN1, effective against drug-resistant gonorrhea and MRSA. This breakthrough addresses the critical global threat of antimicrobial resistance. However, the AI-driven discovery raises complex legal questions regarding the ownership of inventions when artificial intelligence plays a significant role in their creation, particularly in light of recent UK patent law rulings.
In a significant stride against the escalating global health crisis of antimicrobial resistance (AMR), researchers from the Broad Institute of MIT and Harvard, as part of MIT’s Antibiotics-AI Project, have successfully designed two entirely new antibiotic compounds using generative artificial intelligence. This groundbreaking work, detailed in a recent publication in the journal Cell, introduces compounds NG1 and DN1, which have demonstrated potent activity against two high-priority drug-resistant pathogens: Neisseria gonorrhoeae and methicillin-resistant Staphylococcus aureus (MRSA).
The urgency for new antibiotics is underscored by the grim statistic that drug-resistant bacterial infections contribute to nearly 5 million deaths globally each year. Traditional drug discovery methods have struggled to keep pace with evolving bacterial resistance, leading researchers to explore innovative approaches.
Led by James Collins, the Termeer Professor of Medical Engineering and Science at MIT and a senior author of the study, the team employed deep learning models to screen an immense chemical space, computationally evaluating over 36 million to 45 million potential compounds. They utilized two distinct generative AI platforms—Chemically Reasonable Mutations (CReM) and a fragment-based variational autoencoder (F-VAE)—to design molecules with desired antibacterial properties, non-toxicity, and drug-like characteristics. “We’re excited about the new possibilities that this project opens up for antibiotics development,” stated Collins. “Our work shows the power of AI from a drug design standpoint, and enables us to exploit much larger chemical spaces that were previously inaccessible.”
The results were highly promising: NG1 exhibited strong efficacy against drug-resistant N. gonorrhoeae in both cell cultures and a mouse model, while DN1 successfully cleared MRSA infections in a mouse skin infection model. These newly designed molecules are structurally distinct from any existing antibiotics and appear to operate through novel mechanisms, primarily by disrupting bacterial cell membranes.
Beyond the scientific triumph, this AI-driven discovery brings to the forefront complex legal and ethical questions surrounding intellectual property ownership. If an AI system designs a novel compound, who holds the rights to the invention? In the United Kingdom, a 2023 ruling by the UK Supreme Court in the case of Thaler v Comptroller-General of Patents explicitly stated that an AI system cannot be named as an inventor under current law. This ruling creates a nuanced situation for inventions where AI plays a pivotal role.
However, the research team emphasizes the indispensable contribution of human expertise. The paper, which lists over 30 authors including lead authors Aarti Krishnan, Melis Anahtar, and Jacqueline Valeri, highlights that human researchers were crucial in selecting synthetically tractable compounds and refining the AI-generated leads. For instance, the team synthesized approximately 100 analogues of the initial AI-designed molecules, with some demonstrating even greater potency. While AI opened up vast chemical possibilities, human ingenuity was essential in identifying and synthesizing viable drug candidates.
Also Read:
- AI-Generated CRISPR Enzyme Open-Sourced, Advancing Gene Editing Capabilities
- ScienceDirect AI Revolutionizes Scientific Research with Advanced Generative Capabilities
Looking ahead, Phare Bio, a nonprofit antibiotic developer co-founded by James Collins, is actively working on further modifying NG1 and DN1 to prepare them for clinical trials. The team also plans to expand their AI platform to design candidates for other challenging pathogens, including Mycobacterium tuberculosis and Pseudomonas aeruginosa, signaling a new era in the fight against infectious diseases powered by artificial intelligence.
