MIT’s deep learning found an antibiotic for a germ nothing else could kill

Scientists at MIT and Harvard’s Broad Institute and MIT’s CSAIL built a deep learning network that can acquire a broad representation of molecular structure and thereby discover novel antibiotics. The resulting compound, halicin, can destroy a pathogen for which no cure has existed, and it could even help in the fight against coronavirus.

One hundred years ago, the state of the art in finding antibiotics was epitomized by the playful explorations of Alexander Fleming, the Scotsman who discovered penicillin. 

"I play with microbes," Fleming is quoted as having said. "It is very pleasant to break the rules and to be able to find something nobody had thought of."

Today's research in antibiotics is conducted somewhat more mechanically, perhaps, but it's still important to break the rules sometimes, to look where one might not otherwise.

Scientists at the Massachusetts Institute of Technology and Harvard last month described in the scholarly journal Cell how they used a deep learning neural network to identify a molecular compound that's different from most antibiotics. They showed that when the compound is injected in mice, it fights bacteria that no existing drug can eliminate. 

The discovery even has implications for fighting the coronavirus that is causing the Covid-19 disease. 

It's a rule-breaker on many levels: Finding a novel use for an existing compound; using neural nets in place of familiar chemical definitions; and finding an antibiotic that doesn't behave like the usual kind. It's enough to make one believe deep learning forms of AI can change the rules of life sciences. 

The single antibiotic molecule they arrived at, christened "halicin," in honor of the Hal 9000 computer in the movie 2001, is a compound known for years to inhibit the activity of protein kinases that can cause liver damage. It wasn't known that it could serve as an antibiotic, until now. In that way, halicin is characteristic of a recent trend in drug discovery: repurposing, where a known compound finds new uses. 

The training procedure used by MIT scientists first exposed a neural network to over two thousand example molecules and "ground truth" about whether or not they fight E. coli bacteria. The trainer network could then be used to find bacteria-fighting molecules in larger data sets including those with millions of molecules. 

Stokes et al.

What halicin fights is a version of the pathogen known as Acinetobacter baumannii, or A. baumannii, one of the increasing number of "multi-drug resistant" bacteria that can't be fought by existing antibiotics. A. baumannii tends to occur in hospital settings, and can accumulate on all sorts of surfaces, including pillows and bed linen but also blood pressure cuffs. It often attacks critically ill patients and it's been a rising public health problem for years. The MIT authors note that the World Health Organization has marked A. baumannii as "one of the highest priority pathogens against which new antibiotics are urgently required."

The big question, as with most AI-in-medicine efforts, is whether the scientists simply got lucky or if there's a logic to the discovery that points the way to further breakthroughs.

There's much here to suggest the latter, logic rather than luck. The scientists trained a model to form a representation of the chemical structure of molecules, and it was that model that picked out the compound they found, a compound that would ordinarily seem unlikely.

To understand the logic at work, you have to consider the problem the scientists faced, a problem a lot of AI grapples with: exploration versus exploitation, how to expand the search for possible answers, but also build upon what's already known. 

Much of antibiotics research at the moment is experiencing a kind of crisis of exploitation and exploration, according to lead author Jonathan Stokes, who is a Banting Fellow in the Collins Lab at the Broad Institute of MIT & Harvard. Antibiotic research has in recent years either turned up duplicative molecules that don't go beyond existing antibiotics, or else gotten stymied trying to find possibilities in the vast searchable chemical space created by high-throughput screening. 

The solution is to leverage AI. Stokes and the team — it's a big team, as Stokes is joined in the paper by nineteen colleagues from multiple MIT and Harvard labs — trained a neural net on known molecules that do and don't fight the bacteria Escherichia coli. Once the network was trained to classify whether a molecule could fight E. coli, they used that trained network to search a database of over 6,000 molecules that are in various stages of clinical development, to pick out one that would match what the neural net concluded is a structure that can fight E. coli, and there they found halicin. 

Lo and behold, that same E. coli antibiotic, halicin, is also good at fighting other bacteria, in particular A. baumannii, the presence of which it drastically reduced in mice who'd been infected with the bacteria. "The isolate that we used in our skin infection model (A. baumannii CDC 288) was resistant to all antibiotics commonly used to treat this bacterium," Stokes wrote in an email to ZDNet.

(One of the stand-out virtues of this work is that it's not just searching in silico, but also testing in vivo; that kind of thorough "wet lab" work is rare in deep learning research, as Science Magazine's Derek Lowe has pointed out.).

Not only was halicin an unexpected antibiotic, it seems, as far as they can tell, to operate via an "unconventional mechanism" that they had to study to try and understand.