The Antibiotics-AI Project will develop seven new classes of antibiotics to treat seven of the world’s deadliest bacterial pathogens over seven years. Harnessing the power of artificial intelligence and deep learning, the Collins Lab at the Massachusetts Institute of Technology (MIT) will dramatically expand our antibiotic arsenal, helping us fight deadly infections that threaten millions of lives.
Main photo caption: At Collins Lab, work doesn’t just happen in petri dishes, but alsothrough powerful machine-learning models. / Shutterstock
On track to be completed in the summer of 2021, researchers at the Collins Lab will soon have access to a chemical library of ~37,000 molecules. These molecules will help form the vbasis of brand new antibiotics to tackle seven of the most deadly pathogens in the world.
The lab achieved one of the most highly-discussed scientific discoveries of the year — the discovery of halicin, a powerful new drug that can kill many strains of antibiotic-resistant bacteria. Collins Lab’s finding, which was the result of a deep learning algorithm, was published in Cell. This work on halicin, alongside other machine-learning endeavors, has been vital in their quest to increase the rate at which new antibiotics are discovered and developed.
Collins Lab has also applied its breakthrough approaches in the fight against COVID-19: developing the first FDA-approved CRISPR-based diagnostic test for SARS-CoV-2; verifying the role of super-spreading events in transmission; highlighting the limitations of herd immunity in preventing transmission; and repurposing the tuberculosis vaccine BCG as a COVID-19 vaccine candidate.
The Collins Lab has launched several new, innovative projects that harness deep learning, synthetic biology and network biology to address the COVID-19 pandemic.
- The development of a facemask diagnostic for detecting SARS CoV-2.
- The development of the first FDA-approved, CRISPR-based lab diagnostic test for SARS-CoV-2 that was co-developed with Sherlock Biosciences (a spinout from the Collins Lab) and is now being manufactured by Integrated DNA Technologies (IDT).
- Using synthetic biology to repurpose the tuberculosis vaccine BCG as a COVID-19 vaccine candidate.
- Using network biology modeling and physics analyses to reveal that coronavirus superspreading is fat-tailed, indicating that superspreading events play a larger, more problematic role in disease transmission than previously thought.
- Showing that populations with herd immunity are susceptible to exogenous sources of SARS-CoV-2, which can lead to significant, continued increases in viral transmission.
“Despite having limited access to our labs, we were able to adapt our efforts and harness the power of deep learning and synthetic biology to help create novel diagnostics, therapeutics, and vaccine candidates for SARS-CoV-2. This pandemic has also brought to the fore the urgency of harnessing the power of AI to dramatically expand our antimicrobial arsenal.”