The role of AI in the post-antibiotic world: Researchers from IBM have designed AI that can generate the design of molecules for new antibiotics, according to Vox. The AI system can reduce the traditionally expensive and time-intensive process of designing new antimicrobial peptides (AMPs) to a matter of days, and has outperformed other leading methods for synthesizing AMPs, according to the research paper published in Nature.

Why does this matter? Experts have warned of the post-antibiotic world, a phenomenon caused by our over-reliance on amoxil-info.net antibiotics in the treatment of crops, farm animals and humans. Globally, around 700k people die of drug-resistant diseases each year, and the World Health Organization warns this number could rise to 10 mn by 2050, exacerbated by using antibiotics to treat secondary infections during the Covid-19 pandemic.

How the IBM AI works: The system uses a generative model with three basic steps. First, the researchers feed a large database of known peptide molecules to the AI. The model learns the basic rules of molecule design by analyzing the relationship between molecules and their properties. From there, researchers can input their desired and undesirable traits, like low toxicity, and the AI will design the molecule based on the parameters. This method has reportedly successfully designed two antimicrobial peptides with wide ranging applications in safely treating pathogens in lab mice.

AI technology had another groundbreaking success in biology recently, when Alphabet subsidiary DeepMind successfully solved the elusive protein folding problem using their AI system AlphaFold. The system successfully predicted how amino acids will fold to create proteins, an issue that stumped science for over 50 years and has far reaching implications on the manufacturing of new medicine and understanding existing diseases caused by misfolded proteins.