The rise of the robotic scientist
Speakers at EFMC International Symposium on Medicinal Chemistry discusses automation and robotics
Day 3 at EFMC International Symposium on Medicinal Chemistry and the topic that seems to be gripping the conference is the rise of robotics in medicinal chemistry.
Martin Burke (University of Illinois, IL, USA) kicked off the morning’s session with a lively presentation on making small molecule synthesis simpler and automatic. In a bid to access the extraordinary functional potential of natural products, Burke’s group wish to generate a more generalized approach to the synthesis of these complex chemicals that are the bugbear of many synthetic chemists. By looking for common motifs among natural products, such as polyenes (of which there are ~3000 within the Chembl library), Burke has managed to simplify the synthesis of a plethora of natural products via an iterative building block assembly process, synonymous with that used to make peptides. Taking this one step further, Burke has developed a machine that it capable of running these experiments automatically! His next goal is to launch a multidisciplinary project, dubbed the Natural Productome, which will remove the obstacles involved in natural product synthesis, and draw on the expertise of chemical physicists, biochemists, biologists etc. so that we may explore the full medicinal potential of natural products.
Following Burke’s presentation, he then rather excitedly welcomed Ross King (University of Manchester, UK) to the stage, who is the creator of the robot scientists Adam and Eve. The former is the first machine ever to discover novel scientific knowledge independently of its human creators. The robot itself is able to:
- hypothesize to explain observations
- devise experiments to test these hypotheses
- physically run the experiments using lab robotics
- interpret the results from the experiments
- repeat the cycle as required
The second generation robot, Eve, has a number of advantages over her predecessor in that she is capable of automating early-stage drug development; integrating drug screening, hit conformation, and cycles of QSAR hypothesis learning and testing. Initially focussed on neglected tropical diseases, Eve has discovered lead compounds against malaria and African sleeping sickness. The robot’s work now lies in the field of cancer research. King concluded his presentation by stating that “there is no fundamental reason why a robot scientist isn’t better than a human”.
Derek Lowe (Vertex Pharmaceuticals, MA, USA), who has gained notoriety through his blog In the Pipeline, rounded off the day’s events with his presentation on the future of robotics and artificial intelligence in medicinal chemistry. Any medicinal chemists feeling threatened by this point, due to the heralded rise of robotic scientists, will have found very little solace in Lowe’s talk.
Lowe began by taking a look back over the past 100 years or so, drawing on examples where machine had replaced man in scenarios we now take for granted – for example the futuristically named ‘robot helper’, who back in the 1920s was responsible for operating an elevator, and who has now been replaced by the fully automated elevator! Lowe commented that the automation of chemistry is inevitable and actually if a scientist was to look around their lab, they would see that automation has already begun to take hold. The rise of robotics should not be resisted, but embraced, since history proves that machines can perform more efficiently, more accurately and more cheaply than a human. In addition, as machine learning becomes more advanced, the opportunities for robots to innovate is becoming a reality. Lowe concluded by warning that as machines get better, medicinal chemists should concentrate on making sure that they can do the things that machines can’t.
Some excellent food for thought from day 3 of the conference. Follow me @fsgfmc where I will be providing updates from #EFMC_ISMC