Can You Say “Chemputer”?
I learned a new word this week: “chemputer.” It’s not a new word — it’s been around since at least 2012 — but chances are, unless you are a chemist or maybe a synthetic biologist, it’s not a word you knew it either. Even if you don’t care about chemistry, biology, or, for that matter, etymology, this is something you might want to pay attention to, because it may end up revolutionizing healthcare.
The term is credited to Professor Lee Cronin of the University of Glascow. Back in 2012, when he was first discussing the concept, he told The Guardian: “Basically, what Apple did for music, I’d like to do for the discovery and distribution of prescription drugs.”
Fast-forward most of a decade and a pandemic, and Dr. Cronin and others are closing in on that goal — although they’ve updated their analogy to “Spotify for chemistry.”
I won’t pretend to understand either the chemistry nor the programming involved, but, simply put, chemputers automate the production of molecules — including prescription drugs, such as, for example, COVID-fighting Remdesivir. CNBC recently profiled activity in the field, spurred by some new papers from Dr. Cronin and Dr. Nathan Collings of SRI Biosciences.
The new paper from Dr. Cronin and his collaborators appeared in Science earlier this month, with the catchy title A universal system for digitization and automatic execution of the chemical synthesis literature. The big breakthrough is more automation of the process, allowing robotic systems to do most of the work.
Dr. Cronin described their work:
What we’ve managed to do with the development of our ‘Chemical Spotify’ is something similar to ripping a compact disc into an MP3s. We take information stored in a physical format, in this case a scientific paper, and pull out all the data we need to create a digital file which can be played on any system, in this case any robot chemist, including our robotic system which is an order of magnitude lower cost than any other similar robot.
Dr. Cronin’s team uses a chemical description language called XDL. CNBC says: “XDL is to the “chemputer” as HTML is to a browser — it tells the machine what to do.” Software called SynthReader scans descriptions of chemical processes, usually natural language processing (NPL) and translates them into XDL, when the chemputer then can actually execute in the lab. The code can be corrected without programming knowledge and the process is hardware independent.
It’s not entirely free of human involvement — “The human will always need to be there to make sure you don’t have a dumpster on fire.” Dr. Cronin believes — but they are “dedicated to making chemical synthesis accessible to everyone, regardless of training.”
Dr. Cronin has big ambitions:
We’re hoping that the system we’ve built will massively expand the capabilities of robot chemists and allow the creation of a huge database of molecules drawn from hundreds of years’ worth of scientific papers.
Our system, which we’re calling Chemify, can read and run XDL files which have been shared among users. Putting that kind of knowledge directly in the hands of people with access to robot chemists could help doctors make drugs on demand in the future.
He brags: “We’ve invented the CPU [central processing unit] for chemistry. That’s really important right now, because all the chemistry robots in the world are not only expensive, but they can’t be programed in the same way.”
Kim Branson, the global head of artificial intelligence and machine learning at GSK, is wowed, telling CNBC: “The chemputer as a concept and the work [Cronin]’s done is really quite transformational.”
Dr. Collins’ latest research has a similar title — Fully Automated Chemical Synthesis: Toward the Universal Synthesizer — and reports similar breakthroughs. Their synthesizer AutoSyn “makes milligram-to-gram-scale amounts of virtually any drug-like small molecule in a matter of hours.” Their paper demonstrated synthesis of ten known drugs and predicts success for a high percent of many other FDA approved small molecule drugs.
Dr. Collins is a big believer in the combination of AI and automation to improve the pharma R&D process. He wrote earlier this year: “Progress in AI offers the exciting possibility of pairing it with cutting-edge lab automation, essentially automating the entire R&D process from molecular design to synthesis and testing — greatly expediting the drug development process.”
“The majority of chemistry hasn’t changed from the way we’ve been doing it for the last 200 years. It’s very manual, artisan driven process,” Dr. Collins told CNBC. “There’s billions of dollars of opportunity there.” No wonder Dr. Branson and other pharma executives are paying close attention.
Darpa is also paying close attention. SRI International, the parent of Dr. Collins’ Bioscience division, just received $4.3 from Darpa for a tool to help automate production of therapeutics for pandemics and other biological threats.
Darpa also is funding a Make-It program to automate “small molecule discovery and synthesis to propel the field beyond conventional batch-based, intuition-driven capabilities,” and a related Accelerated Molecular Discovery program, in which, as Anne Fisher, the program manager, told CNBC: “We’re now trying now to harness what we’ve done in Make-It and expand it out so we can teach computers how to discover new molecules.”
Think about that “Teach computers how to discover new molecules” and let that sink in. As Dr. Collins says, “This is still a very new science. It’s started to explode really in the last 18 months.”
All this is taking place as 3D printing for pharmaceuticals is also starting to take off, such as for “low-cost production of customized pill medications for patients who need special dosing, quantities or composition of drugs. Pills can be 3D printed in unique sizes, shapes and with slow-release capabilities.” The FDA is still working on how to regulate 3D printing of medical products (which now include prostheses, orthopedic and other implants, pharmaceuticals, and even organs).
It better start thinking about chemputers, or at least the products made by them.
At the very least, we can expect that chemputers and 3D printing could greatly speed and democratize the production of pharmaceuticals. Imagine your doctor or pharmacist producing your medicine on the spot — or perhaps doing it yourself, in your own home. Further development of AI could also greatly speed up on the discovery process, which could have major implications not just for our health but also for the pharmaceutical industry. The old models are up for grabs.
So, get to know chemputers. They may be in your future.
Please follow me on Medium and on Twitter (@kimbbellard), and don’t forget to share if you liked the article!