Artificial Intelligence is Transforming Drug Discovery

Alind Vats
Nov 25 · 3 min read

Artificial Intelligence expedites and optimizes the drug discovery process. Though approx US$2.8 billion is spent on drug development annually, most of it is wasted as only 1 out of 10 candidate therapies make it past regulatory approval. The money can be spent more effectively by utilizing computational power and intelligent systems.

Capitalizing on these capabilities, Pfizer is using IBM Watson, a system that uses machine learning, to power its search for immunooncology drugs. Sanofi is using Exscientia’s AI platform to identify metabolic-disease remedies. Genetech uses an AI system from GNS Healthcare to drive the search for cancer treatments.

A biotech company, Berg, has developed a model to identify previously unknown cancer mechanisms using tests on more than 1,000 cancerous and healthy human cell samples. They modelled diseased human cells by varying the levels of sugar and oxygen the cells were exposed to, and then tracked their lipid, metabolite, enzyme and protein profiles.

The group uses its AI platform to generate and analyse immense amounts of biological data from patients to highlight key differences between diseased and healthy cells. This approach tremendously largens the training data that informs novel drug discovery. The company also uses this AI system to look for drug targets and therapies for diabetes and Parkinson’s disease.

Berg aims to identify potential treatments on the basis of the precise biological causes of disease. “We are turning the drug-discovery paradigm upside down by using patient-driven biology and data to derive more-predictive hypotheses, rather than the traditional trial-and-error approach,” says a co-founder.

BenevolentBio, a London startup, has created an AI platform that consolidates data from research papers, patents, clinical trials and patient records. The data is then used to establish representations of biological entities such as genes, symptoms, diseases, proteins, tissues, species and candidate drugs, and to infer how they relate.

Example of a knowledge graph linking entities

The generated knowledge graph can be queried like a search engine to study medical conditions, their associated genes, and compounds affecting the conditions. AI is going to lead to a better understanding of human biology and will give us the means to fully address human diseases. AI will be to biology what mathematics is to physics.

AI has the potential to pin-point previously unknown causes of diseases. This will accelerate personalized medication tailored to fit the biological profile of the patient. Diseases will become more preventable and treatable.

Robots are used in conjunction with AI systems to expand drug discovery. A robot called Adam was the first to discover new scientific knowledge by identifying the function of a yeast gene. Adam scoured public databases and generated hypotheses about which genes code for key enzymes that catalyse reactions in the yeast Saccharomyces cerevisiae. Adam then physically tested its predictions in a lab.

Researchers then independently tested Adam’s hypotheses about the functions of 19 genes; 9 were new and accurate, and only 1 was wrong. Later, a more advanced robot, Eve, discovered that triclosan, a common ingredient in toothpaste, could potentially treat drug-resistant malaria parasites.

PhD programs will evolve to reflect the innovations and demands. They will combine the understanding of human biology with computer science, computational statistics and statistical machine learning. Though deep research into novel territories in a specification, like a particular gene mutation, is the norm right now, the PhD academic curricula will be broader a decade from now.