AI is transforming the healthcare industry and has the potential to disrupt and improve the way we discover drugs and will allow us to drastically improve the current methods of biomarker discovery. This week, Insilico Medicine announced in their paper Converging blockchain and next-generation artificial intelligence technologies to decentralize and accelerate biomedical research and healthcare their most recent breakthroughs, bringing together AI with blockchain technologies to give you access to your own medical data to ensure that you have total control and ownership of your own data to ‘redefine how pharma is done’.

Insilico Medicine, a Baltimore U.S based company, are bringing to market a ‘secure and transparent distributed personal data marketplace utilizing blockchain and deep learning technologies may be able to resolve the challenges faced by the regulators and return the control over personal data including medical records back to the individuals.’

Why is now the time to introduce DL into drug discovery pipelines?

Deep learning is currently powering our autonomous cars, transforming art, the way we manufacture, sample, and play video games, and this is the technology that is going to revolutionise drug discovery.

Alex Zhavoronkov, CEO at Insilico Medicine explained how ‘DL has surpasses human accuracy in many tasks for example image descriptions, which Andrej Kaparthy from Tesla really opened our eyes about, and there’s also super human accuracy in playing all kinds of games.’

This is being employed in healthcare in the following ways:

  • Dermatologist-level classification of skin cancer with deep neural networks
  • Autonomous driving - this 4 year old technology is being rolled out and passengers are being transported from point A to point B at 100 mph, one mistake and you’re dead, yet we’re doing it - so why aren’t we doing it in pharma?
  • Governments are competing to be ahead of the game and to take leadership in AI and DL. Countries are opening up their roads so that they don’t fall behind on autonomous driving, and pharma is similarly going to be transformed - in the US the FDA is looking into this issue and one of the countries leading the way is Korea where they’re looking at the legislations to find new ways to accelerate the pharmaceutical drug development process using AI.

In the coming years pharma regulations will change, and for good reason: out of all industries, pharma is one of the most inefficient and the productivity is going down. It currently costs more than $2.6bn to develop a molecule into a drug, and there’s a 92% failure rate. Now, with the increased availability of data and recent advancements in AI there are countless opportunities for transforming the industry with DL to provide the next generation of innovative pharmaceutical solutions that may be used to accelerate the biomedical research and enable patients with new tools to control and profit from their personal data as well with the incentives to undergo constant health monitoring.

Upon speaking with Polina Mamoshina, Senior Research Scientist and lead contributor of the paper, she said:

Most people do not understand what life data they have, how valuable and dangerous this data may be and do not have any control over how their life data is being used. The policy makers are trying to address this problem by introducing new regulations that make it expensive and difficult for the innovators to turn the human life data into life-saving products. In this paper we propose a blockchain-enabled solution to help people become aware of and take control over their data and profit from licensing the data to the innovators.

There’s an abundance of public data available, and between 1990 and 2010, over $20bn in public funds was spent on high-throughput experiments alone. You can now generate much more data through those experiments per dollar, and Insillico are training their models on this. They have introduced for the first time half-life period of analysis significance, models of data value for single and group of users and the cost of buying data in the context of biomedical applications.

Upon speaking with Lucy Ojomoko, Blockchain AI program manager at Insilico Medicine, she explained the following:

The proposed data value model will not only allow to transform Big Data into a new type of data - Apprised Data, but also will play a role in the development of data collection culture, and will lay a new approach to data analysis, revealing the overall significance of data to the certain health condition of the patient.

If you’re keen to learn more about Insilico Medicine’s most recent progressions, Polina will be attending the Women in AI in Healthcare Dinner next week in London (Nov. 22).