In the future, the medicines we take may be "made by AI"

In the future, the medicines we take may be "made by AI" [Artificial intelligence is participating in the whole process from target discovery to clinical trials] 　　The birth of a new drug usually requires an investment of 1 billion or even billions of dollars, and the research and development cycle generally exceeds 10 years. Due to the addition of AI technology, the cost of drug research and dev

In the future, the medicines we take may be "made by AI" [Artificial intelligence is participating in the whole process from target discovery to clinical trials]

　　The birth of a new drug usually requires an investment of 1 billion or even billions of dollars, and the research and development cycle generally exceeds 10 years.

Due to the addition of AI technology, the cost of drug research and development today has been significantly reduced, and the research and development time has also been greatly shortened.

For example, AI reduced the time to preclinical compound candidates from an average of 4 1/2 years to about 13.7 months, a reduction of nearly 75%.

　　◎Our reporter Chen Xi

　　Recently, according to media reports, the team of Professor David Baker of the University of Washington published a paper in the journal Cell, using an artificial intelligence (AI) technology platform to accurately design a macrocyclic polypeptide molecule that can pass through the cell membrane from scratch, opening up the design A new approach to new oral medicines.

　　In recent years, AI has accelerated the development of new drugs, and has almost participated in the entire process from drug target discovery to clinical trials.

During the COVID-19 epidemic, AI is behind the advent of many drugs, and the global AI pharmaceutical industry has accelerated.

　　AI is integrated into all aspects of drug development

　　"The term AI was coined by John McCarthy at the 1956 Dartmouth Conference to describe 'the science and engineering of making intelligent machines'. AI was also introduced into the field of drug discovery around this time." Nankai University According to Lin Jianping, a professor at the School of Pharmacy, in 1964, the establishment of the field of quantitative structure-activity relationship modeling became a sign that AI began to be used in drug research and development.

　　Today, AI is playing an increasingly important role in drug development and is closely integrated with all aspects of drug development.

　　It takes a long and bumpy process to develop a drug from scratch.

It mainly includes 4 stages of research and development, namely target selection and validation, compound screening and lead optimization, preclinical research and clinical trials.

And each stage involves many specific links.

　　For example, Lin Jianping said, for example, in the stage of target selection and validation, disease-related targets need to be identified.

Determining targets based on traditional experiments is time-consuming and costly. Using AI technology combined with existing omics big data, based on known and newly generated experimental data, potential candidate targets can be quickly analyzed, saving time and energy. cost; or when the efficacy of the lead compound is known, but the specific mechanism of action and side effects are unclear due to the lack of a clear target, AI can predict the target on a large scale, narrow the range of candidate targets, and finally combine experimental methods to quickly locate the real target.

"AI helps drug developers to quickly find targets and accelerate the process of transforming lead compounds into drugs." Lin Jianping introduced.

　　For existing drugs, AI can also discover new targets through target prediction, thereby discovering new drug indications, which is also a very hot field - drug repositioning.

　　In the most important clinical trial stage, the application of AI has also played a multiplier effect.

"At this stage, it is necessary to evaluate the safety and effectiveness of drugs in patients, and AI can participate in patient recruitment, clinical trial design, and data analysis of trial results." Lin Jianping gave an example. Among patients, data such as personal characteristics, symptoms, and treatment effects of patients are extracted to find the patients that best match the current trial; in terms of trial design, AI can predict appropriate drug doses, treatment plans, etc.; in terms of trial data, AI technology can be used to track And manage the real-time situation of patients, predict patient prognosis, etc.

　　AI greatly reduces drug development costs

　　The birth of a new drug usually requires an investment of 1 billion or even several billion US dollars. The research and development cycle generally exceeds 10 years, but the success rate is less than 10%.

Due to the addition of AI, today's drug development costs have been reduced by hundreds of millions of dollars, and the development time has also been greatly shortened, generally by more than half.

For example, AI has shortened the development time of preclinical candidate compounds from an average of 4.5 years to about 13.7 months, a reduction of nearly 75%.

　　In addition, AI has also improved the success rate of drug development.

"In layman's terms, drug research and development is actually a trial-and-error process. AI can help us eliminate a lot of mistakes, leaving us with a greater chance of success in the end," said Lin Jianping.

　　It is precisely because AI pharmaceuticals have the advantage of rolling over traditional pharmaceuticals that the AI pharmaceutical industry has grown globally.

At present, the development of AI pharmaceutical industry can be summarized into three stages: in the first stage, AI pharmaceutical companies are initially formed, mainly providing AI technical services for drug research and development at a certain stage; in the second stage, AI pharmaceutical companies have developed mature R&D Pipeline, and the developed drugs enter clinical validation, this stage will attract a large amount of capital and start-ups to join; and the third stage, enter the key clinical phase II efficacy experiments, truly prove the effectiveness of AI-developed drugs.

　　"At present, the global AI pharmaceutical industry has entered the third stage of development." Lin Jianping said.

　　AI pharmacy in my country started late and is still in the second stage.

"But the domestic AI pharmaceutical industry is developing very fast. Major Internet giants and some large pharmaceutical companies have begun to deploy AI pharmaceuticals, including some start-ups, of course." Lin Jianping said.

　　According to statistics, there are currently more than 60 AI pharmaceutical companies in China. Last year, the financing scale of AI pharmaceuticals in my country reached 1.236 billion US dollars, a year-on-year increase of 163.54%.

　　There are many challenges in AI pharmacy

　　It can be said that AI has penetrated into all aspects of the field of drug research and development, promoting the upgrading of the pharmaceutical industry, and is very likely to bring about changes in the pharmaceutical industry in the future.

With the current development of the AI pharmaceutical industry, in the near future, we may soon usher in the first innovative drug developed by AI technology.

In addition to their expectations, many people also have doubts about whether the drugs developed by AI are risky.

　　"At present, the risks of drugs we develop using AI are the same as those of traditional drug development, including drug side effects, toxicity, tolerance, etc." Lin Jianping explained that because AI currently plays a major role in drug development The auxiliary effect still needs to be verified by real tests to verify its safety and effectiveness. Experts are required to evaluate it, so the risk is the same as that of traditional R&D drugs.

But doing so also brings another problem. The pharmaceutical industry is still based on expert experience, which has become the biggest obstacle to the development of AI pharmaceuticals.

"The reason for this phenomenon is mainly due to the distrust of AI technology to help pharmaceuticals." Lin Jianping believes that with the successful launch of AI drugs in the next few years, this problem will definitely be solved; on the other hand, AI is currently in In the whole process of drug research and development, it still plays the role of an auxiliary tool and does not occupy a dominant position, which also determines that it is difficult for the AI pharmaceutical industry to achieve rapid development.

　　Moreover, AI technology is still developing, and breakthroughs in data, algorithms, and computing power will also take a certain amount of time.

Such as insufficient data volume, uneven data quality, low algorithm accuracy, and inability to meet the needs of the algorithm, all of which have brought difficulties to the development and application of AI in drug development.

　　In addition, AI pharma faces many other challenges.

For example, there are still many unsolved problems in basic theoretical research in the field of life; another example is the lack of compound talents, "Those who understand computing do not understand pharmacy, and those who understand pharmacy do not understand computing", how to better transform biological problems into computational problems , and then use digital means to solve it, which requires the participation of a large number of compound talents, and the training of such talents is also extremely time-consuming.

In the future, the medicines we take may be "made by AI"

You may like

Trends 24h

Latest