Power of AI: Only 21 Days from Target Discovery to Candidate Molecule!

Posted by Lisa George on November 27th, 2020

The past decade has witnessed an year-by-year increase in the number of patients diagnosed with cancer and diabetes worldwide, and at the same time, drug research and development is carrying out intensively. According to data reported, the global pharmaceutical R&D investment in 2016 was US5.4 billion (excluding pharmaceutical production investment), and the compound annual growth rate from 2012 to 2016 was 2.4%. In the context of rising new drug R&D costs, it is estimated that global pharmaceutical R&D investment will reach 160 billion US dollars in 2021, and the compound annual growth rate will reach 1.9% from 2016 to 2021.

Although the global enthusiasm for new drug research and development is high, it is not easy to make commendable achievements. It is estimated that the successful launch of a new drug often carries an average R&D cycle of more than 10 years and a R&D cost of more than 2 billion USD. In addition, the research and development process of a new drug consumes a lot of energy, and the risk of R&D failure is relatively high. According to a research report released by Deloitte, the return on investment of the world’s top 12 biopharmaceutical giants in R&D in 2017 was only 3.2%, the lowest ever in the past eight years. However, the R&D cost of a new drug on the market has increased from 1.188 billion US dollars to 2 billion US dollars within eight years.

Fortunately, with the advent of the era of artificial intelligence, AI has opened up new doors for new drug development. At present, AI is gradually coming into the sight of pharmaceutical industry, especially in the research and development of new drugs.

Experts said that AI technology is something that comes to rescue, which will change the long process of traditional drug research and development, such as developing drug targets, leading compound screening, preclinical animal experiments, and clinical trials, and greatly improving the success rate of drug development. In the process of traditional drug development, in order to find suitable candidate molecules, thousands of small molecules need to be tested, and assuming that a few candidate molecules are found, the next level of screening is required. In the end, only ten molecules can enter clinical trials.

At present, global pharmaceutical giants are making efforts in this regard, allowing AI to participate in drug research and development, and quickly analyze the results of drug structure, disease pathophysiological mechanisms, the efficacy of existing drugs, and sample observation under the microscope.

For example, as early as 2016, Johnson & Johnson in the United States sold some small molecule compounds that were still in trials to an artificial intelligence company, hoping to accelerate the development of new drugs. In 2017, GlaxoSmithKline of the United Kingdom cooperated with artificial intelligence startups to use big data and machine learning to accelerate the development of innovative small molecule drugs. In 2018, Novo Nordisk announced the reorganization of its R&D center to enhance its competitiveness in AI and accelerate the expansion and diversification of its severe chronic disease product pipeline.

In September, 2019, the results of the research and development of new AI drugs were published in Nature Biotechnology. Scientists at InsilicoMedicine, WuXi AppTec, and the University of Toronto have adopted a new AI system called Generative Tension Reinforcement Learning (GENTRL) to shorten the time for new drug development from target to candidate molecule to only 21 days. Compared with the time needed in the past few months or past years, this number is undoubtedly a big surprise! This method also marks a breakthrough in AI-assited medicine at the industrial level. In addition, InsilicoMedicine, known as the top 100 artificial intelligence companies in the world in 2018, also has a strategic cooperation with WuXi AppTec.

Scientists from AI & Medicine has successfully developed a "AI-powered drug discovery" platform for pharmaceutical companies that are seeking for AI assistance. The platform provides a broad and integrated portfolio of medical and scientific solutions in areas like drug R&D, medical translation, medical imaging, medical therapy and research system, and more. To sum up, application of AI into drug discovery has become a global trend in the pharmaceutical industry. Although it can only play an supporting role at present, it has indefinite potentials in the future with further exploration being carried out.