[Reporter's view] AI drug development, should face reality

"Artificial intelligence (AI) drug development will be useful, but there is a problem with only talking too much hope.

 

Domestic AI drug development is still at an early stage, so more R&D is needed." These are the words the representative of an AI drug development company attending a recently held event related to digital healthcare threw to the debaters in the audience.

 

It seems to point out that although new drug development using AI is being carried out in various parts of the world, companies that have achieved tangible results in Korea have yet to be seen.

 

According to the Ministry of Science and ICT, new drug development is one of the projects that take an average of 10 years or more for a new drug to come to the market.

 

It takes about 5 years to discover 5,000 to 10,000 promising candidates.

 

Among the discovered candidate substances, 10 to 25 substances are entering preclinical trials.

 

Of these, nine substances are approved for phase 1 clinical trial plans.

 

The success rate is so low that there are only 5 in Phase 2, 2 in Phase 3, and only 1 in the market.

 

In order to overcome the low success rate of new drug development and entry barriers that require long-term and high costs, a plan to utilize AI in the early R&D stage has been proposed.

 

For traditional new drug development, a method of selecting a target disease and searching for a candidate substance by a research team reviewing about 400 to 500 related papers is used.

 

AI is expected to have the advantage of being able to search for more than 1 million papers and hundreds of thousands of chemicals so that dozens of researchers can do the work that they have to do for 1 to 5 years more quickly.

 

The AI drug development market is estimated to have high growth potential.

 

The AI drug development market is expected to grow from 636.2 billion won in 2019 to 4.7693 trillion won by 2027, growing at an average annual rate of 28.63%.

 

Global pharmaceutical companies such as AbbVie, Amgen, AstraZeneca, Boehringer Ingelheim, GSK, Janssen, Eli Lilly, MSD, Novartis, Pfizer, Sanofi, and Roche have also actively collaborated with AI drug developers to discover candidate substances.

 

In 2021, a candidate substance for the treatment of pulmonary fibrosis discovered and designed through AI by Insilico Medicine in Hong Kong entered clinical trials.

 

In October of last year, Verge Genomics, a US AI drug developer, started clinical trials of candidate materials for amyotrophic lateral sclerosis (ALS, Lou Gehrig), which were secured using AI's new drug discovery platform Converge.

 

Verge Genomics is a biotech company that has received investments from Eli Lilly, MSD, WuXi AppTec, and BlackRock.

 

In Korea, the Ministry of Science and ICT invested 25.8 billion won for three years from 2019 to 2021 to develop new AI drugs.

 

The main goal was to build a customized AI platform for each stage of new drug development.

 

It aimed to lead the development of three areas: candidate substance discovery, drug re-creation, and smart pharmacovigilance.

 

Through a three-year support project in the field of AI new drug development, it developed an AI model that can be applied to neurodegenerative diseases and new anticancer drugs.

 

The developed AI model is loaded into the public platform KAIDD and is openly operated so that researchers from industry, academia, and research institutes can utilize it.

 

As a follow-up project that started in September of last year, the company plans to improve the data sharing and utilization environment by advancing the AI new drug development model by 2026.

 

We plan to develop new drug candidates that can be applied for a clinical trial plan.

 

There are several areas that need improvement in clinical prediction and targeting of clinical patient groups.

 

It is also an important issue to establish social consensus and specific action plans that can utilize personal medical data classified as sensitive information.

 

It is problematic that the AI drug development industry is not securing tangible results such as sales while talking only about the ideal use of technology, overseas cases, and the growth potential of the global AI market.

 

It is necessary to face reality and solidify internal stability to be competitive.