- x NeonPulse | Future Blueprint
- Posts
- 🤖 Machine learning and the future of pharma
Good morning and welcome to the Sunday edition of neonpulse!
Today’s email is going to focus on one of the most interesting use cases of A.I. and machine learning: new drug development.
Machine learning and the future of pharma
One of the more interesting realities of pharmaceutical drugs is that not all drugs work for all people, and one of the places that this is most apparent in cancer treatment.
Due to genetic variability, different cancer fighting drugs may be effective for one person and completely ineffective for another, which can lead to a prolonged and painful course of treatment.
In traditional chemotherapy, doctors work their way down a list of common cancer drugs, testing their patients response until they find one that works.
Exscientia’s stated mission is to “combine the power of AI and human creativity to make safer and more sophisticated drugs available to all,” and the company has already begun harnessing the power of machine learning to create a revolutionary drug matchmaking protocol.
The goal?
To make the “trial and error” drug selection process a thing of the past.
By taking a small amount of tissue from patients, researchers are able to split the sample into more than 100 pieces, exposing the samples to various types of drugs simultaneously.
Computer vision powered by machine-learning is then used to analyze the effectiveness of each of the drugs, significantly speeding up the drug selection process and avoiding months of potentially ineffective chemotherapy.
One patient recently profiled in an excellent MIT technology review article underwent this new therapy after 6 courses of failed chemotherapy:
“The approach allowed the team to carry out an exhaustive search for the right drug. Some of the medicines didn’t kill the cancer cells. Others harmed his healthy cells. The patient was too frail to take the drug that came out on top. So he was given the runner-up in the matchmaking process: a cancer drug marketed by the pharma giant Johnson & Johnson that the patients doctors had not tried because previous trials had suggested it was not effective at treating his type of cancer.”
And it worked.
Two years post-treatment, the patient’s cancer was in complete remission, showcasing the incredible power of the new drug selection technology.
Yet Exscientia is looking to solve more than just drug selection with A.I. - they’re also working to apply machine learning to new drug development.
Creating new drugs is an incredibly expensive and time consuming process, on average taking more than 10 years and billions of dollars to develop an FDA approved product, but the goal of Exscientia is to cut both cost and development time down significantly with the help of A.I.
“The whole process of drug discovery is about failure,” says biologist Richard Law, chief business officer at Exscientia. “The reason that the cost of coming up with a drug is so high is because you have to design and test 20 drugs to get one to work.”
Yet by using A.I. to predict how drugs might behave in the body, researchers are able to discard dead-end compounds early in the process, allowing them to focus their time on more promising candidates.
And with how lucrative blockbuster drugs can be for pharmaceutical companies, there are now hundreds of startups looking to harness the power of machine learning to accelerate the drug development process.
The progress these companies have made in a short period of time is impressive, with over 20 A.I. designed drugs already in the clinical trial stages.
But beyond simply developing new drugs with existing compounds, there’s another promising area that machine learning and A.I. can help drug researchers- helping to find new biological and chemical structures that can be used to create future drugs.
According to Adityo Prakash, CEO of the California-based drug company Verseon, there are billions upon billions of molecules that could be used to make new drugs, but the pharmaceutical industry only currently works with around 10 million of these molecules…
Meaning that we’re barely scratching the surface of whats possible.
“We’re not even fishing in a tide pool next to the ocean,” said Prakash, “We’re fishing in a droplet.”
You can read more in the MIT Technology Review article here.
And now your moment of zen
That’s all for today folks!
If you’re enjoying Neon Pulse, we would really appreciate it if you would consider sharing our newsletter with a friend by sending them this link: