Multiple therapeutic areas to benefit from AI drug discovery collaboration

Date: 11th September 2019

One of the greatest challenges during the drug discovery process is identifying drug candidates that are both safe and effective. Whilst these parameters have been addressed by testing thousands of compounds, determining targets, monitoring off target effects, and increasing the potency of candidates, this optimisation is resource intense and time-consuming.  The introduction of artificial intelligence (AI) into this field has, however, the potential to revolutionise drug discovery and design.  AIs’ ability to analyse large datasets, recognise patterns, learn and adapt has already been utilised to increase the speed at which a small number of highly specific drug candidates can be synthesised and tested.

One company, San Francisco-based Atomwise, was ahead of the curve in seeing the value of AI for drug discovery and they were one of the first companies to deploy AI in this manner.  Their technology uses a statistical approach extracting the insights from millions of experimental affinity measurements and thousands of protein structures to enable the prediction of small molecules binding to proteins, based on convolutional neural networks similar to those recognising faces in crowds or in self-driving cars.  The announcement of an AI-drug discovery partnership with Hansoh Pharma worth potentially up to $1.5B will benefit multiple therapeutic areas, with the aim to design and discover potential drug candidates for up to eleven undisclosed target proteins.  Hansoh Pharmaceutical Group (Hansoh Pharma) is a leading pharmaceutical company in China dedicated to discovering and developing life-changing medicines and this collaboration is hoped to produce first in-class and best in-class therapies. We will be watching with interest what comes next…

For more information please read the press release