Deep learning algorithm could boost drug development

Artificially intelligent algorithms can learn to identify amazingly subtle information, enabling them to distinguish between people in photos or to screen medical images as well as a doctor. But in most cases their ability to perform such feats relies on training that involves thousands to trillions of data points. This means artificial intelligence doesn't work all that well in situations where there is very little data, such as drug development.

Vijay Pande, professor of chemistry at Stanford University, and his students thought that a fairly new kind of deep learning, called one-shot learning, that requires only a small number of data points might be a solution to that low-data problem.

"We're trying to use machine learning, especially deep learning, for the early stage of drug design," said Pande. "The issue is, once you have thousands of examples in drug design, you probably already have a successful drug."

The group admitted the idea of applying one-shot learning to drug design problems was farfetched -- the data was likely too limited. However, they'd had success in the past with machine learning methods requiring only hundreds of data points, and they had data available to test the one-shot approach. It seemed worth a try.

Much to their surprise, their results, published April 3 in ACS Central Science, show that one-shot learning methods have potential as a helpful tool for drug development and other areas of chemistry research.

Moving from images to molecules
Other researchers have successfully applied one-shot learning to image recognition and genomics, but applying it to problems relevant to drug development is a bit different. Whereas pixels and bases are fairly natural types of data to feed into an algorithm, properties of small molecules aren't.

To make molecular information more digestible, the researchers first represented each molecule in terms of the connections between atoms (what a mathematician would call a graph). This step highlighted intrinsic properties of the chemical in a form that an algorithm could process.

With these graphical representations, the group trained an algorithm on two different datasets - one with information about the toxicity of different chemicals and another that detailed side effects of approved medicines. From the first dataset, they trained the algorithm on six chemicals and had it make predictions about the toxicity of the other three. Using the second dataset, they trained it to associate drugs with side effects in 21 tasks, testing it on six more.

In both cases, the algorithm was better able to predict toxicity or side effects than would have been possible by chance.

"We worked on some prototype algorithms and found that, given a few data points, they were able to make predictions that were pretty accurate," said Bharath Ramsundar, who is a graduate student in the Pande lab and co-lead author of the study.

However, Ramsundar cautioned that this isn't a "magical" technique. It was built off of several recent advances in a particular style of one-shot learning and it works by relying on the closeness of different molecules, as indirectly indicated by their formula. For example, when the researchers trained their algorithm on the toxicity data and tested it on the side effect data, the algorithm completely collapsed.

An experimentalist's help
People concerned about AI taking jobs from humans have nothing to fear from this work. The researchers envision this as groundwork for a potential tool for chemists who are early in their research and trying to choose which molecule to pursue from a set of promising candidates.

"Right now, people make this kind of choice by hunch," Ramsundar said. "This might be a nice compliment to that: an experimentalist's helper."

Beyond giving insight into drug design, this tool would be broadly applicable to molecular chemistry. Already, the Pande lab is testing these methods on different chemical compositions for solar cells. They have also made all of the code they used for the experiment open source, available as part of the DeepChem library.

"This paper is the first time that one-shot has been applied to this space and it's exciting to see the field of machine learning move so quickly," Pande said. "This is not the end of this journey - it's the beginning."

This research was funded by the Fannie and John Hertz Foundation.

Han Altae-Tran, Bharath Ramsundar, Aneesh S. Pappu, Vijay Pande.
Low Data Drug Discovery with One-Shot Learning.
ACS Cent. Sci., doi: 10.1021/acscentsci.6b00367.

Most Popular Now

GSK reaches agreement with Novartis to acquire ful…

GlaxoSmithKline plc (LSE/NYSE: GSK) today announces that it has reached an agreement with Novartis for the buyout of Novartis' 36.5% stake in their Consumer Healthcare Jo...

Canadian neuroscientists say daily ibuprofen can p…

A Vancouver-based research team led by Canada's most cited neuroscientist, Dr. Patrick McGeer, has successfully carried out studies suggesting that, if started early enou...

Merck partners with Medisafe to help improve medic…

Merck, a leading science and technology company, today announced a new collaboration with US-based Medisafe to help its cardiometabolic patients better manage medication ...

New immunotherapy for lung cancer shows promise of…

In a groundbreaking development, results from a recent clinical trial to treat lung cancer show that a novel immunotherapy combination is surprisingly effective at contro...

Boehringer Ingelheim and OSE Immunotherapeutics an…

Boehringer Ingelheim and OSE Immunotherapeutics, a biotechnology company focused on the development of innovative immunotherapies, have announced a collaboration and excl...

Personalized tumor vaccine shows promise in pilot …

A new type of cancer vaccine has yielded promising results in an initial clinical trial conducted at the Perelman School of Medicine at the University of Pennsylvania and...

Taking a standard prostate cancer drug with food b…

By taking a high-cost drug with a low-fat meal - instead of on an empty stomach, as prescribed - prostate cancer patients could decrease their daily dose, prevent digesti...

New targeted therapy schedule could keep melanoma …

Skin melanoma, a particularly insidious cancer, accounts for the vast majority skin cancer deaths and is one of the most common cancers in people under 30. Treatment for ...

Researchers identify chemical compound that inhibi…

An organic chemical compound shows effective antiviral activity against Ebola virus and several other viruses, according to a study led by Georgia State University. The r...

When drugs are wrong, skipped or make you sick: Th…

Rising drug prices have gotten a lot of attention lately, but the actual cost of prescription medications is more than just the dollars and cents on the bill. Researchers...

New class of drugs could help tackle treatment-res…

Researchers have discovered a new class of drug that has the potential to help cancer patients who no longer respond to existing therapies. The drug may not become availa...

Researchers propose key elements of antimicrobial …

Antimicrobial stewardship programs (ASPs) in hospitals play a vital role in managing the threat of antibiotic resistance. To be of maximum effectiveness, essential elemen...