Daewoong Pharmaceutical digested 800 million compounds to facilitate AI drug discovery

Publicly listed Daewoong Pharmaceutical has recently unveiled its AI-driven drug development system, powered by a database of hundreds of millions of pre-processed compounds. 

WHAT IT’S ABOUT

Over the past two years, researchers of the Seoul-based pharma giant have sought to solve perennial issues in the development of novel drugs, particularly high cost and low efficiency. 

Serving as the foundation of its latest drug discovery system, DAVID (Daewoong Advanced Virtual Database) is an 800 million-compound database composed of recently discovered compounds and compounds collected from research on new drugs over the past 40 years. 

The company said they had to preprocess these open-source compounds, filtering out unusable information, as not all are suitable for AI-driven drug development. Daewoong claims there are potentially an estimated one billion new drug candidates on DAVID.

After building DAVID, Daewoong was able to develop AIVS (AI-based Virtual Screening), a tool used for uncovering active substances for target proteins. It runs on generative AI that can quickly discover new patentable active substances for target proteins, as well as search these substances in various ways based on 3D modelling. 

Utilising both DAVID and AIVS, Daewoong built its AI-based system for novel drug development, Daisy (Daewoong AI System). Introduced last year, the web-based system allows Daewoong researchers to discover new compounds and quickly predict drug properties. It also enables ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) research. 

WHY IT MATTERS

Through Daisy, Daewoong was able to uncover an active substance that reacts to two target proteins related to weight loss and treatment of diabetes. It was also able to discover within six months an effective active substance for cancer cell inhibition, securing a lead substance, which would have taken between one to two years to find using existing methods.

It has also claimed success in the development of protein degraders, “dramatically” reducing research trial and error through simultaneous conducts of antibody design and safety evaluation, and ultimately shortening the time for new drug development using AI. 

Soon, the local pharma giant plans to expand the use of its AI throughout the entire drug development cycle, including preclinical, clinical, and commercialisation stages.

MARKET SNAPSHOT

Daewoong named DAVID after the Biblical hero who defeated Goliath, symbolising its mission to compete with global big pharma in the race to develop novel drugs using AI. 

Last year saw some of the world’s pharma giants, Boehringer Ingelheim and AstraZeneca, get into drug discovery partnerships that will harness generative AI. Pfizer also tied up with Tempus for cancer drug development. 

Citing data from Korea Health Industry Development Institute, Daewoong said at best, it takes an average of 15 years to develop new drugs with only one out of 10,000 candidate substances usually becoming a success. 

A recent report by the Korea Food and Drug Safety Evaluation Institute suggested that by applying AI, this process can be cut down to half (or seven years), reducing costs down by 600 billion won (around $450 million). 

ON THE RECORD

“The world of new drug candidates is like the universe, and it is no exaggeration to say that AI has opened a new era of exploration in drug development. If we break ground in unknown areas with AI, we can discover a large number of new drug candidates. We will develop excellent new drugs more quickly and make a significant contribution to human health,” said Park Jun-Seok, director of the New Drug Discovery Center at Daewoong.