Medication

AI and robotics join forces to revolutionize the way medicine is made – Advanced Science News

Scientists are accelerating the production of drugs to breathe new life into old medicines and reduce the risk of failure in clinical trials.

AI, machine learning, and robotics have greatly improved drug discovery and manufacturing. But in the expensive and risky pharmaceutical business, these tools still have room to make an impact in an overlooked part of the development pipeline: drug production.

After drug discovery and before clinical testing and manufacturing,​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​ with a tablet, nasal spray or sample.

Although not so important, this complex part of the development of any new medicine is often overlooked in this developing technology.

According to researchers at the University of Toronto, with the help of AI, the automated methods of drug production are changing and offer untapped opportunities, which in addition to producing new drugs that work well, can restore the old ones.

Filtering through a multitude of possibilities

If the drug is the carrier, drug design is the design of the plane, says Christine Allen, professor and researcher in the Department of Pharmacy at the University of Toronto. Although many drugs stick to the usual methods that have been used before, there is little guarantee that this is the best.

He explained: For an oral medicine, he says there are more than 10 billion possible combinations of ingredients or ingredients, “and no one is going to test those 10 billion possibilities.” .”

AI and machine learning algorithms will prove to be the perfect tool to sift through these many possibilities. However, an early attempt to use AI in Allen’s lab to predict how a drug would be released from a polymer layer revealed flaws in this theory.

“We took data from the literature, took some from our lab, and used that data to train machine learning methods,” Allen said. It wasn’t long before problems arose.

“There’s missing information or missing information in the literature, and that’s the information you’re going to need to train machine learning methods,” he said.

Contacting robots

To help fill in the gaps, Allen’s team began using automated lab equipment to quickly perform routine tests and generate missing data quickly and accurately.

The team had already brought two powerful tools into the lab – AI and automation – but it wasn’t until they let go of conventional big ideas and combined the two in a new way that their ability to make drugs that were fulfilled.

The first revelation was moving away from large training datasets. Instead, Allen and colleagues created an AI-powered robotics platform. He explained: “We don’t really rely on these big data tools. “What we do is connect the drug that we like. We plug in all the different tools or tools that we can use when we’re done, and then we’re able to develop from a series of goals.”

These factors are everything from the size of the tablet to how quickly the tablet breaks down, to how much drug is released, and even the cost of materials. Out of the billions of possible forms, the AI ​​chooses a suitable starting set and begins to adapt to the correct one.

The automated platform quickly creates and evaluates the first forms after which the data is fed back to train the algorithm and generate the next set of forms. “It’s a closed loop until we know the options that fit the product profile,” Allen said.

Allen believes that this approach has the potential to not only speed up research, but also increase efficiency. According to him, better drugs can help more drugs than in clinical trials.

“This is what creates many of the medical failures we see, the use of best practices,” he said. In this way, drugs that enter the medical center can be designed to give them a better chance of success.

Breathing new life into pharma

This approach can also revive promising molecules that have failed clinical trials and improve existing drug formulations. Allen’s startup, Intrepid Labs, does just that.

“What the company is doing is looking at drugs that are off patent, or coming off patent, and we’re repurposing them,” Allen said.

In tests carried out in his laboratory, they made drugs days to weeks longer than those currently sold for properties such as bioavailability, which is the amount of a drug that makes it into the bloodstream.

The presence of AI and automation in the areas of drug discovery and manufacturing should facilitate their integration into the pharmaceutical company’s workflow.

Allen and his team have been paying attention to scalability while developing this system. “I see this as a very easy transition,” he said.

“We need to stop accepting this very high failure rate associated with drug development, and we just need to think about accepting and embracing these technological advances throughout the entire pipeline.”

Reference: Christine Allen, et al., The Dawn of a New Pharmaceutical Epoch: Can AI and Robotics Transform Drug Manufacturing? Advanced Healthcare Devices (2024). DOI: 10.1002/adhm.202401312

#robotics #join #forces #revolutionize #medicine #Advanced #Science #News

Leave a Reply

Your email address will not be published. Required fields are marked *