The increase in the number of superbugs with antibiotic resistance has prompted tech company IBM to develop artificial intelligence that can invent new antibiotics, according to IBM Research Blog. That system has created two new potential medications.
With the rate of bacteria developing resistance to antibiotics, experts project that infections currently considered minor could take 10 million lives per year come 2050, noted New Atlas. This is aggravated by the fact that new and effective antibiotics take a long time to develop.
IBM seeks to address this issue with the help of an AI system that helps explore the possible chemical combinations that can fight infections. It uses the deep generative autoencoder model to look at a wide range of sequences and captures details regarding their functions.
The model also takes a look at the peptides’ molecule configurations and other similar peptides.
After examining peptides, the AI uses the Controlled Latent Attribute Space Sampling (CLaSS), which uses the data gathered in the first step. Then, the AI creates new peptides that contain specific attributes, particularly antimicrobial properties.
IBM explained, “We performed additional screening with the help of high-throughput, coarse-grained molecular dynamics simulations.”
“These simulations look for presence of novel physiochemical features indicative of stable and peptide-membrane binding, such as low-contact variance between peptide and membrane.”
IBM’s artificial intelligence system was able to create 20 antimicrobial peptides in less than two months. Upon studying the novel candidates, two of them showed potential in fighting Gram-positive and Gram-negative pathogens.
The two peptides were also able to fight multidrug-resistant K. pneumoniae and are unlikely to create drug resistance in E. coli.
“We also didn’t find any cross-resistance for either of the AMPs when tested using a polymyxin-resistant strain. Live-cell confocal imaging showed the formation of membrane pores as the underlying mechanism of bactericidal mode of action of these peptides,” said the IBM blog.
The researchers tested the two antimicrobials in vitro and on mice, which showed low toxicity. This is an indication of safety and efficacy in a complex animal system.
The tech company proposes the use of its artificial intelligence to quickly and efficiently create “potent and selective broad-spectrum antimicrobials to keep antibiotic-resistant bacteria at bay – for good.”
The company is also optimistic that its development will be able to contribute to the battle against bacteria and infections they cause in an environmentally friendly and sustainable manner.