close
close

Nvidia-backed artificial intelligence company Iambic announces ‘breakthrough’ in drug development

Nvidia-backed artificial intelligence company Iambic announces ‘breakthrough’ in drug development

Martin Coulter

LONDON (Reuters) – Biotech firm Iambic Therapeutics unveiled on Tuesday what it says is a revolutionary artificial intelligence model that could radically reduce the time and money needed to develop new drugs.

An increasing number of technology startups are using AI to advance pharmaceutical research. Iambic, which previously received investment from tech giant Nvidia, has released details of its new artificial intelligence drug development model called “Enchant.”

Enchant was trained on a large amount of preclinical data obtained from laboratory tests of drugs conducted before they were tested in humans. The model was developed to predict how a particular drug will work at the earliest stage of development.

In a white paper published by Iambic, Enchant demonstrated a high degree of accuracy in predicting how well the human body would metabolize certain drugs, with the results compared to real-world results.

The company said its model set a new standard with a prediction accuracy score of 0.74. By comparison, earlier models only achieved 0.58.

Iambic co-founder and CTO Fred Manby told Reuters that researchers using Enchant could potentially halve the investment required to develop some pharmaceuticals because they could see how successful a drug could be at a very early stage.

“The cost of bringing a product to market is often estimated to be around $2 billion, and much of this has to do not with program costs but with failure rates. The cost of getting a product to market is due to the high likelihood of late-stage failure,” he said.

“If you achieve a 10% improvement at each stage of clinical development, you will essentially cut the cost in half because it is applied cumulatively.”

Frances Arnold, winner of the 2018 Nobel Prize in Chemistry and a member of Iambic’s board of directors, told Reuters the development represents a major advance in the use of AI for drug discovery.

Referring to Google’s DeepMind AlphaFold program, whose developers recently won the Nobel Prize in Chemistry, Arnold said Enchant solves a different problem in the drug discovery process.

“AlphaFold predicts the three-dimensional structure of how a molecule binds to a target protein, but structure is not enough,” she said.

“The success of a drug candidate is determined by its pharmacokinetic, efficacy and toxicity properties. Enchant addresses these distinct and important issues.”

(Reporting by Martin Coulter; Editing by Susan Fenton)