In February 2020, a team of researchers at MIT leveraged artificial intelligence (AI) to identify an antibiotic capable of combatting E. coli and an antibiotic-resistant strain of Acinetobacter baumannii. Drawing inspiration from the Stanley Kubrick movie, “2001: A Space Odyssey,” about a rogue AI, “HAL,” they named their new drug “halicin.”
That research illustrated the remarkable speed of AI-assisted drug discovery. Scientists trained their AI model using around 2,500 molecules, a mix of FDA-approved drugs, natural products, and more. Once trained to recognize molecules effective against E. coli, the AI screened 6,000 compounds, including existing drugs, natural products, and others.
Halicin was identified significantly faster than traditional methods, marking a major advancement in addressing antibiotic-resistant “superbugs,” which is an growing public health concern.
By identifying structural components and functional groups within a molecule, scientists can determine its functions. In May 2023, a separate research team published their research, building upon the methods used in the original halicin discovery, which identified another superbug-fighting antibiotic candidate.
These discoveries showcase the speed and potential of AI in drug discovery, which can swiftly identify powerful new antibiotics from thousands of molecules. AI’s power lies in its ability to analyze vast medical data, paving the way for rapid drug development, potentially improving patient outcomes, and reducing costs.
The role of AI in medicine has evolved over time, with a recent focus on deep learning and generative learning, which is revolutionizing drug discovery. Advancements such as DeepMind’s Alphafold, the cataloging of biological and chemical data, and AI’s predictive capabilities have and will further accelerated drug discovery.
AI is a valuable tool but cannot replace human involvement in drug development, particularly in the later formulation, testing and trial stages. Yes, questions about AI’s reliability in safety-critical medical decisions remain, but the potential to transform drug discovery is both established and accelerating.
The ainewsarticles.com article you just read is a brief synopsis; the original article can be found here: Read the Full Article…
