AI antibiotics: How artificial intelligence algorithms are identifying new types of antibiotics

• Author:
• Author name

  Quantumrun Foresight
• April 19, 2022

Insight summary

The exploration of new processes to discover antibiotics, such as incorporating artificial intelligence (AI), is opening doors to combat highly resistant bacteria. This research has far-reaching implications, from potentially saving millions of lives to shaping government policies and influencing the pharmaceutical industry's direction. Additional discoveries, such as solutions to protein folding and the development of carbon-absorbing materials, are poised to make significant impacts on healthcare, climate change mitigation, and energy production and storage.

AI Antibiotics context

Researchers using AI models at the Massachusetts Institute of Technology (MIT) identified (2020) an antibiotic that can kill many bacteria strains. Critically, this drug (Halicin) can kill strains of bacteria that are known to be resistant to antibiotics. By using AI models, antibiotics can be identified by only indicating the properties of the desired molecule, compared to initiating countless trials and tests that could take years to deliver a result and at considerable cost. MIT researchers used a computer model capable of screening more than a hundred million chemical compounds in days. The model is designed to select potential antibiotics that kill bacteria using different mechanisms than those used by other drug treatments. 

US-based technology firm IBM has also developed an AI system that can design new antibiotics in days and help create new treatments for other conditions such as the COVID-19 pandemic. The company's system relies upon a generative model, with researchers providing the AI algorithm with a massive database of known peptide molecules. The AI algorithm uses the database to analyze patterns and then calculate the relationship between molecules and their properties.

If a particular molecule is found to have a certain structure or composition, it tends to perform a specific function, which allows the system to learn the basic rules of molecule design. Researchers can tell the AI system what properties it wants the new molecule to have, and using these parameters, then designs new molecules that meet these parameters, which can then be tested on mice. The AI system is designed to scan up to 100 million chemical compounds within a few days and identify prospective antibiotics.

Disruptive impact

The discovery of new antibiotics through a process that scientists are exploring represents a beacon of hope in the fight against antibiotic-resistant bacteria. This process not only allows for the identification of antibiotics that can target highly resistant variants but also opens the door to designing new drugs that can combat bacteria no longer affected by known antibiotics. As we enter a 'post-antibiotic' era, where existing antibiotics are losing their effectiveness due to overuse in treating plants, animals, and humans, this research is more critical than ever. 

According to a United Nations report, the reduced efficiency of antibiotics is a serious concern, leading to up to 700,000 deaths annually. This number is projected to rise to an alarming 10 million by 2050 if the trend continues. The ability to discover and create new antibiotics is not just a scientific advancement; it's a lifeline that could potentially save millions of lives in the future. For governments, this underscores the importance of supporting research and development in this area, as well as implementing policies to regulate the overuse of existing antibiotics. 

The development of new antibiotics may lead to more effective treatments for common infections, reducing the risk of complications and death. Companies in the pharmaceutical industry may need to invest in this research, recognizing its potential to shape the future of medicine. Educational institutions may also need to focus on training the next generation of scientists in this field, ensuring that the momentum continues. 

Implications of AI antibiotics

Wider implications of AI antibiotics may include:  

• Pharmaceutical applications developing new treatments and drugs to treat previously untreatable diseases, leading to extended life expectancy and improved quality of life for patients suffering from chronic and acute conditions.


• Possible solutions being discovered to the decades-long problem of protein folding, enabling a deeper understanding of biological processes and accelerating the development of therapies for diseases like Alzheimer's and Parkinson's.


• The discovery of carbon-absorbing materials to tackle climate change, contributing to a reduction in greenhouse gas emissions and supporting global efforts to mitigate the effects of global warming.


• The discovery of materials that can be used to advance the production and storage of energy, enhancing energy efficiency and reducing dependence on non-renewable resources.


• Governments implementing stricter regulations on antibiotic usage in agriculture and healthcare, leading to a more controlled and responsible approach to preserving the effectiveness of existing antibiotics.


• Increased nvestment in research related to antibiotic resistance, fostering collaboration between scientists, healthcare professionals, and policymakers, and driving the development of new medical technologies.


• The development of targeted educational campaigns to raise awareness about antibiotic resistance, leading to more informed public behavior and support for responsible antibiotic usage.


• A shift in the pharmaceutical industry's focus towards personalized medicine, leveraging the understanding of individual genetic makeup, leading to more tailored and effective treatments.


• The integration of new materials and technologies into construction and manufacturing, leading to more energy-efficient buildings and transportation systems, contributing to a more sustainable urban environment.


• The emergence of new job roles and specializations within the scientific and healthcare sectors, leading to a demand for skilled professionals in fields related to antibiotic research, protein folding, and material science.

Questions to consider

• What do you think may limit the ability of pharmaceutical AI systems to produce new antibiotics, if any? 


• Do you believe that any new antibiotics discovered as part of AI-driven research efforts should be made readily available to countries that would not have the resources to produce and perform similar research?