Chemists have a new lab assistant: artificial intelligence. Researchers have developed a ‘deep learning’ computer program that produces blueprints for the sequences of reactions needed to create small organic molecules, such as drug compounds. The pathways that the tool suggests look just as good on paper as those devised by human chemists.
The tool, described in Nature on 28 March1, is not the first software to wield artificial intelligence (AI) instead of human skill and intuition. Yet chemists hail the development as a milestone, saying that it could speed up the process of drug discovery and make organic chemistry more efficient.
“What we have seen here is that this kind of artificial intelligence can capture this expert knowledge,” says Pablo Carbonell, who designs synthesis-predicting tools at the University of Manchester, UK, and was not involved in the work. He describes the effort as “a landmark paper”.
Chemists have conventionally scoured lists of reactions recorded by others, and drawn on their own intuition to work out a step-by-step pathway to make a particular compound. They usually work backwards, starting with the molecule they want to create and then analysing which readily available reagents and sequences of reactions could be used to synthesize it — a process known as retrosynthesis, which can take hours or even days of planning.
The new AI tool, developed by Marwin Segler, an organic chemist and artificial-intelligence researcher at the University of Münster in Germany, and his colleagues, uses deep-learning neural networks to imbibe essentially all known single-step organic-chemistry reactions — about 12.4 million of them. This enables it to predict the chemical reactions that can be used in any single step. The tool repeatedly applies these neural networks in planning a multi-step synthesis, deconstructing the desired molecule until it ends up with the available starting reagents.
Segler and his team tested the pathways that the program threw up in a double-blind trial, to see whether experienced chemists could tell the AI’s synthesis pathways from those devised by humans. They showed 45 organic chemists from two institutes in China and Germany potential synthesis routes for nine molecules: one pathway suggested by the system, and another devised by humans. The chemists had no preference for which was best.
Researchers have been trying to use computing power to plan organic chemical synthesis since the 1960s, with only limited success. But Segler’s tool is one of several programs developed in recent years that use AI to flag up potential reaction routes.
Chematica, the most well-known, was acquired by German pharmaceutical company Merck in May 2017, for an undisclosed sum. Bartosz Grzybowski, a chemist at the Ulsan National Institute of Science and Technology in South Korea, and his team spent years inputting the rules of organic chemistry into the system for the program to draw on.
Earlier this month, Grzybowski reported2 that he had tested eight of his algorithm’s suggested pathways in the laboratory, and that they all worked. “I’m very glad there is this revival of retrosynthesis, and welcome different approaches,” he says.
Segler’s tool is different because it learns from the data alone and does not need humans to input rules for it to use.
Ola Engkvist, a computational chemist at pharmaceutical company AstraZeneca in Gothenburg, Sweden, is impressed by the work. “Increasing the success rate in synthetic chemistry would have a huge benefit in terms of speed and efficiency on drug-discovery projects, as well as cost reduction,” he says.
Segler says his tool has already piqued the interest of several pharmaceutical companies. But he doesn’t see it putting organic chemists out of work. “It will be an assistant for the chemist who wants to make molecules and get from A to B as quickly as possible,” he says. “The GPS navigation device may render paper maps redundant — but not the driver of the car.”
Read the related News & Views article: ‘AI designs organic syntheses’
Segler, M. H. S., Preuss, M. & Waller, M. P. Nature 555, 604–610 (2018).
Klucznik, T. et al. Chem 4, 522–532 (2018).