Artificial intelligence (AI) is increasingly being utilised in the drug discovery process, with the latest breakthrough being the development of an antibiotic that is powerful enough to combat resistant bacteria strains via a deep learning algorithm.
AI has the power to enable biopharmaceutical companies to reduce costs and improve success rates in drug discovery and development, according to stakeholders and investment experts. Composed of machine learning and deep neural networks, AI is a technology that is perfectly suited to combing through masses of medical data, find candidate molecules and predict therapeutic targets.
Until recently, AI had yet to enter the realm of antibiotic development. The World Health Organisation (WHO) states that antimicrobial resistance (AMR) “threatens the effective prevention and treatment of an ever-increasing range of infections caused by bacteria, parasites, viruses and fungi” and is “an increasingly serious threat to global public health that requires action across all government sectors and society.” (Source)
The need for effective antibiotics continues to increase, as without them, healthcare costs will continue to rise, illnesses may become prolonged, and major surgeries and treatments could be compromised. Furthermore, drug-resistant bacteria could put as many as 10 million lives at risk by the year 2050 without the appropriate antibiotics to target those bacteria.
In a groundbreaking study, AI has been employed by researchers at the Massachusetts Institute of Technology (MIT) to help fill the need for new antibiotics. The researchers trained a deep learning algorithm by feeding the program data about specific features of over 2,000 drugs and compounds in order to identify the types of molecules that can kill bacteria.
After teaching the algorithm about important molecular features, the researchers used it to analyse a digital library of over 6,000 compounds that differed from existing antibiotics, hoping these new potential drugs could effectively kill bacteria before the bugs had a chance to develop resistance. One of the compounds, given the name “halicin” by the research team, promised to be particularly powerful in killing resistant strains.
The research team hopes to utilise the same algorithm to find antibiotics that are more selective in the bacteria they kill, according to senior researcher Regina Barzilay. This would mean a significant improvement in the way antibiotics work – targeting only the bugs that cause an infection, and not all the healthy bacteria that live in the gut. The team also plans to use the algorithm to design powerful new antibiotics. (Source)
