How deepfakes and AI are being used to find new ways to treat diseases

18 Nov 2022

Image: © Jacob Lund/Stock.adobe.com

Drug discovery companies such as Insilico Medicine are using deepfake AI technology to design new molecules that can help treat diseases.

Click to read more stories from Future Health Week.

Intel made a splash earlier this week when it unveiled its latest technology that can detect a deepfake in real-time with 96pc accuracy. AI such as this can help organisations around the world to prevent the spread of misinformation and protect themselves from cybercrime.

But not all kinds of deepfakes are bad. The advancement of any emerging technology brings with it positive and negative uses – and the future of healthcare certainly has much to gain from deepfakes.

On Monday (14 November), clinical-stage drug discovery company Insilico Medicine unveiled a spate of updates to its Pharma.AI SaaS platform, showcasing advancements the company has made in its biotech business by harnessing the power of artificial intelligence.

One such application of AI is in using deepfake technology, most often used to create lifelike videos of humans, to design new molecules that have the potential to treat diseases.

As an end-to-end AI platform, Pharma.AI can discover new targets, design new drugs and even predict the outcomes of clinical trials. Insilico expects the technology to substantially improve the time and cost it takes to bring new life-saving drugs to the market.

“We are at a tipping point,” said Dr Petrina Kamya, head of AI platforms at Insilico Medicine, who spoke at the launch event earlier this week. The launch comes eight years after rigorous R&D with a team of drug hunters and AI scientists trying to improve healthcare.

“We’ve shown just how far this technology can go, with the first AI-discovered target and AI-designed drug in phase 1 clinical trials. The further we advance, the more futile it becomes to argue against the adoption of AI.”

Speaking to SiliconRepublic.com, Kamya said that the launch was driven by Insilico’s latest addition to the Pharma.AI platform, InClinico – a tool built for clinical trial outcomes prediction and clinical trial landscape monitoring.

“One of the most exciting features of InClinico is the ability to score your trial. This allows you to prioritise your programmes based on their likelihood of success in addition to other metrics,” she said.

While deepfake technology cannot play a direct role in treating diseases, Kamya said that “it can play an important role in generating data where it is lacking or in generating ideas in terms of novel chemistry”.

Rapid discovery of new molecules

Dually headquartered in Hong Kong and New York, Insilico was founded in 2014 by CEO Alex Zhavoronkov. The private biotech company is on a mission to accelerate drug discovery and development by continuously inventing and deploying new AI technologies.

Zhavoronkov told Forbes recently that while big pharma companies all have their own AI teams, the size of the organisations and it takes time to deploy the technology can work against them. This is where Insilico comes in.

It provides top pharma and biotech companies with AI services, such as deepfake technology through the Pharma.AI platform, to enable streamlined R&D efforts and transform the way therapeutics and materials are discovered.

This rapid discovery and development of novel molecules can help in creating therapeutic programmes across a range of health areas including cancer, fibrosis, anti-infectives and immunology.

Just this month, Insilico announced a research partnership with Paris-based pharma giant Sanofi in a deal potentially worth up to $1.2bn to boost the drug discovery efforts of the Sanofi Institute for Biomedical Research, Sanofi’s R&D centre in China.

The use of deepfakes in medicine isn’t just confined to designing novel molecules. Deepfakes can also help hospitals create artificial patients using patient data. This is then used for testing and experimentation that’s realistic – but doesn’t endanger the lives of actual patients.

This means that instead of using real patient data, which raises ethical concerns, researchers and hospitals can use ‘true-to-life’ deepfake patients to test new methods of diagnosis and monitoring – and even train AI to help in making medical decisions as assistants to doctors.

10 things you need to know direct to your inbox every weekday. Sign up for the Daily Brief, Silicon Republic’s digest of essential sci-tech news.

Vish Gain was a journalist with Silicon Republic

editorial@siliconrepublic.com