How AI robots hunt new drugs for crippling nerve disease
Send a link to a friend
[August 10, 2017]
By Ben Hirschler
LONDON (Reuters) - Artificial intelligence
robots are turbo-charging the race to find new drugs for the crippling
nerve disorder ALS, or motor neurone disease.
The condition, also known as Lou Gehrig's disease, attacks and kills
nerve cells controlling muscles, leading to weakness, paralysis and,
ultimately, respiratory failure.
There are only two drugs approved by the U.S. Food and Drug
Administration to slow the progression of ALS (amyotrophic lateral
sclerosis), one available since 1995 and the other approved just this
year. About 140,000 new cases are diagnosed a year globally and there is
no cure for the disease, famously suffered by cosmologist Stephen
"Many doctors call it the worst disease in medicine and the unmet need
is huge," said Richard Mead of the Sheffield Institute of Translational
Neuroscience, who has found artificial intelligence (AI) is already
speeding up his work.
Such robots - complex software run through powerful computers - work as
tireless and unbiased super-researchers. They analyze huge chemical,
biological and medical databases, alongside reams of scientific papers,
far quicker than humanly possible, throwing up new biological targets
and potential drugs.
One candidate proposed by AI machines recently produced promising
results in preventing the death of motor neurone cells and delaying
disease onset in preclinical tests in Sheffield.
Mead, who aims to present the work at a medical meeting in December, is
now assessing plans for clinical trials.
He and his team in northern England are not the only ones waking up to
the ability of AI to elucidate the complexities of ALS.
In Arizona, the Barrow Neurological Institute last December found five
new genes linked to ALS by using IBM's <IBM.N> Watson supercomputer.
Without the machine, researchers estimate the discovery would have taken
years rather than only a few months.
Mead believes ALS is ripe for AI and machine-learning because of the
rapid expansion in genetic information about the condition and the fact
there are good test-tube and animal models to evaluate drug candidates.
That is good news for ALS patients seeking better treatment options.
Famous sufferers include Lou Gehrig, the 1923-39 New York Yankees
baseball player; actor and playwright Sam Shepard, who died last month;
and Hawking, a rare example of someone living for decades with the
If the research goes on to deliver new medicines, it would mark a
notable victory for AI in drug discovery, bolstering the prospects of a
growing batch of start-up companies focused on the technology.
Those firms are based on the premise that while AI robots won't replace
scientists and clinicians, they should save time and money by finding
drug leads several times faster than conventional processes.
Mead from Sheffield is working with BenevolentAI, one of a handful of
British "unicorns" - private companies with a market value above $1
billion, in this case $1.7 billion - which is rapidly expanding
operations at its offices in central London.
[to top of second column]
Physicist Stephen Hawking sits on stage during an announcement of
the Breakthrough Starshot initiative with investor Yuri Milner in
New York April 12, 2016. REUTERS/Lucas Jackson/File Photo
Others in the field include Scotland's Exscientia and U.S.-based firms Berg,
Numerate, twoXAR, Atomwise and InSilico Medicine - the last of which recently
launched a drug discovery platform geared specifically to ALS.
"What we are trying to do is find relationships that will give us new targets in
disease," said Jackie Hunter, a former drug hunter at GlaxoSmithKline (GSK)
who now heads Benevolent's pharma business.
"We can do things so much more dynamically and be really responsive to what
essentially the information is telling us."
Unlike humans, who may have pet theories, AI scans through data and generates
hypotheses in an unbiased way.
Conventional drug discovery remains a hit-and-miss affair and Hunter believes
the 50 percent failure rates seen for experimental compounds in mid- and
late-stage clinical trials due to lack of efficacy is unsustainable, forcing a
shift to AI.
A key test will come with a Phase IIb study by Benevolent to assess a previously
unsuccessful compound from Johnson & Johnson in a new disease area - this time
for treating Parkinson's disease patients with excessive daytime sleepiness.
Big pharmaceutical companies like GSK, Sanofi and Merck are now exploring the
potential of AI through deals with start-ups.
They are treading cautiously, given the failure of "high throughput screening"
in the early 2000s to improve efficiency by using robots to test millions of
compounds. Yet AI's ability to learn on the job means things may be different
CPR Asset Management fund manager Vafa Ahmadi, for one, believes it is a
"Using artificial intelligence is going to really accelerate the way we produce
much better targeted molecules. It could have a dramatic impact on productivity,
which in turn could have a major impact on the valuation of pharmaceutical
stocks," he said.
Drugmakers and start-ups are not the only ones chasing that value. Technology
giants including Microsoft <MSFT.O>, IBM and Google's parent Alphabet <GOOGL.O>
are also setting up life sciences units to explore drug R&D.
For Benevolent's Hunter, today's attempts to find new drugs for ALS and other
difficult diseases marks an important test-bed for the future of AI, which is
already being deployed in other high-tech areas such as autonomous cars.
"The aim is to show that we can deliver in a very difficult and complex area. I
believe if you can do it in drug discovery and development, you can show the
power of AI anywhere."
(Reporting by Ben Hirschler; Editing by Pravin Char)
[© 2017 Thomson Reuters. All rights
Copyright 2017 Reuters. All rights reserved. This material may not be published,
broadcast, rewritten or redistributed.