This browser is not actively supported anymore. For the best passle experience, we strongly recommend you upgrade your browser.


Powered by Bird & Bird

| 1 minute read

What the replication crisis means for AI and drug discovery

Back in May 2016, the scientific publication Nature surveyed over 1,500 researchers to find out how reproducible their work was. This question goes to the very core of the scientific method, and has formed the backbone of countless clinical trials. As it turns out, the results were pretty bleak. More than 70% of those surveyed tried and failed to reproduce the results of another scientist, and 73% said that only around half of the papers in their field can be trusted. This phenomenon, colloquially known as the ‘replication crisis’, poses a real problem for researchers.

Why does the replication crisis matter now more than ever?

Legal practitioners will know that it’s impossible to navigate the press without coming across an article about artificial intelligence (AI). Evangelists are quick to extol the virtues of AI and the role it will play in the future of drug discovery and life sciences generally. The proposition is relatively simple on paper. First, AI-powered algorithms dredge huge datasets from theoretical models, pre-clinical and clinical studies, PhD theses, scientific journals, and everything in between. Therapeutic targets are then identified alongside a suitable drug. The advantages are clearly very tempting. McKinsey recently estimated that almost 270 companies are working in the AI-driven drug discovery industry, 15% of which have an asset in preclinical development.

Ultimately, the AI algorithms used in drug discovery are only as good as the information we feed them. We may also face some interesting legal hurdles. For example, how can we protect confidential information and other IP rights from a seemingly omnipotent algorithm? What about safeguarding patient data? How much weight will regulatory authorities place on AI-generated data with respect to marketing authorisation of novel therapeutics? This list is clearly not exhaustive, but highlights some interesting challenges. Time will tell whether AI systems will rule the lab, but the drug discovery industry is optimistic.

Quoted in a recent article, Jim Weatherall (VP of data science, AI and data science at AstraZeneca) said that AI is the future of drug development. The mindset has changed from “what is this?” to “why did we ever do it any other way?”.

McKinsey recently estimated that almost 270 companies are working in the AI-driven drug discovery industry, 15% of which have an asset in preclinical development.


ai, artificial intelligence, data science, drug discovery, biotech, pharmaceuticals, regulatory