AI Pressuring Pharmaceutical Sector

DeepMind the AI arm of Google’s parent company Alphabet solved a problem that long puzzled scientists at an annual biology conference as it innovated how drugmakers create and develop new medicines.

It started in December 2018 at the CASP13 meeting on Mexico’s northeastern stretch of Caribbean coastline when DeepMind beat seasoned biologists at predicting the shapes of proteins which are the basic building blocks of disease. This may seem like a niche advancement but proteins are instrumental in synthesising new drugs and so an accurate tool that can accurately model protein structures could speed up the development of new drugs. 

Absolutely stunning. It was a total surprise
— University of Maryland Computational Biologist John Moult

Currently in the Pharmaceutical industry solving the structure of proteins in order to find ways for medicines to attack disease is a very complicated process. This is because researchers still don’t fully understand the mechanisms for how proteins are built. Furthermore, there’s a mathematical aspect as there are more possible protein shapes than there are atoms in the universe. Meaning that for the previous 25 years, computational biologists have laboured to devise software equal to the task.

Despite these obstacles in the CASP13 conference DeepMind armed with the latest neural-network algorithms was able to achieve more than what 50 top labs from around the world could accomplish combined. Even though it had limited experience in protein folding which is the physical process by which a protein acquires its three-dimensional shape.

DeepMind Analysing Protein Folding Process

DeepMind Analysing Protein Folding Process

AI could usher in a new era of medical progress
— The Guardian

However, although DeepMind has made many pioneering strides its simulation is currently quite limited because it doesn’t yet produce the kind of atomic-level resolution that is necessary for drug discovery. Additionally, even though many companies are looking for ways to use computers to identify new medications, few AI based drugs have progressed to the point of being tested in humans.

However, there is definitely a large gap for AI to fill in the future as finding new drugs and bringing them to market is notoriously difficult. With some estimates suggesting big drugmakers spend more than $2.5 billion to get a new medicine to patients. As a result, many leading Pharmaceutical Companies such as Recursion are prioritising rapid implementation of AI having recently raised $1.08 billion from Venture Capitalists over four times more than last year at just $237 million.

If we want to understand the other 97 per cent of human biology, we will have to acknowledge it is too complex for humans
— Chris Gibson, the CEO of Recursion Pharmaceuticals.

Inspired by Alphabet many other large firms, which are centred in other sectors, are expanding into the Pharmaceutical industry with the help of AI. A notable example is Facebook which released a paper in April 2019 using deep learning to analyse 250 million protein sequences allowing them to creep into pharmaceutical-company turf.

Mohammed AlQuraishi, a Harvard computational biology researcher sums up the current stance of the Pharmaceutical industry perfectly:

DeepMind beat everyone by a sizeable margin, if drugmakers don’t take the threat seriously they could be left in the dust.
— Mohammed AlQuraishi computational Biologist