Current location - Education and Training Encyclopedia - Graduation thesis - A paper on new drugs.
A paper on new drugs.
The development of new drugs has always been a time-consuming and expensive complex project, but now, scientists seem to have found a solution-introducing artificial intelligence technology.

According to the MIT Science and Technology Review reported on September 3rd, a team of Insilicon Medicine, an artificial intelligence pharmaceutical startup, cooperated with scientists from the University of Toronto to develop new targeted drugs in only 46 days, and completed the preliminary biological verification. The results of this study have been published in this week's Nature Biotechnology.

This landmark research confirms that AI technology can help accelerate drug development, which means that the patent protection period is extended, thus improving the economy of drug development. If this method can be popularized, it will be widely adopted by the pharmaceutical industry.

AI helps to shorten the drug development time from 8 years to 46 days.

Based on two popular artificial intelligence technologies, the team introduced a new tension reinforcement learning (GENTRL) generated by artificial intelligence system in this drug development.

The researchers chose DDR 1 kinase (a tyrosine kinase expressed in epithelial cells) as the target, which is a protein closely related to tissue fibrosis. After determining the target, GENTRL system designed 30,000 different molecular structures within 2 1 day, and then screened out new molecular structures that can be synthesized in the laboratory by consulting the known molecules acting on drug targets in previous studies and patents.

The research and development achievement entitled "Deep learning can quickly identify effective DDR 1 kinase inhibitors" has been published in the journal Nature Biotechnology. Screenshot from Nature Biotechnology magazine

Among the six candidate DDR 1 inhibitor compounds designed and synthesized by GENTRL, four compounds have activity in biochemical analysis. In the next phase of in vitro cell experiment, two of the four active compounds showed the expected inhibition ability of DDR 1 and could effectively reduce the content of markers related to fibrosis process. By comparison, the most potential 1 compound has been further successfully verified in mouse experiments.

From initial target determination, molecular structure screening, synthesis of potential new drugs to preclinical biological verification, GENTRL system shortens the traditional drug research and development method from at least 8 years to only 46 days.

Michael levitt, winner of the Nobel Prize in Chemistry in 20 13 and a professor of structural biology at Stanford University, commented, "This paper is undoubtedly an impressive progress, and it is likely to be applicable to many other problems in drug design. Based on the most advanced reinforcement learning, I am also impressed by the breadth of this research because it involves molecular modeling, affinity measurement and animal research. "

Artificial intelligence is becoming the mainstream to replace the role of pharmaceutical chemists.

MIT Science and Technology Review magazine pointed out that this landmark research may change the dilemma of "wasting money, time and energy" faced by new drug research and development.

This landmark research may change the dilemma faced by new drug research and development. According to MIT science and technology review magazine

"Artificial intelligence will have a revolutionary impact on the pharmaceutical industry, and we need more experimental results to accelerate this progress," said Jürgen Schmidhuber, a professor at the Swiss Institute of Artificial Intelligence, who is the inventor of many core technologies and initial concepts in the field of artificial intelligence.

As we all know, it takes a lot of money and time to bring a new drug to market. According to the data of Tufts Center for Drug Development and Research, it may take 10 years for a new drug to go on the market, costing as much as $2.6 billion, and most candidate drugs will fail in the testing stage.

Therefore, reducing R&D cycle and economic cost is very important for the success of pharmaceutical R&D activities in the pharmaceutical field. According to Forbes magazine, the research and development cost of this drug is only $6,543.8+$5,000 by using the method of Insilicon Medicine.

Insilicon Medicine hopes to bring AI deep learning into the drug research and development process. According to Forbes magazine

Charles Cantor, the chief scientist of the Human Genome Project of the US Department of Energy and a professor at Boston University, said that there are many exaggerated claims about the prospect of artificial intelligence AI in improving medical care and developing new medical tools. However, this achievement recently published in Nature Biotechnology is really significant.

It first proves that artificial intelligence can replace the role usually played by pharmaceutical chemists, and this role is often understaffed; Secondly, the acceleration of drug development means the extension of patent protection period, thus improving the economy of drug development. "If this method can be popularized, it will be widely used in the pharmaceutical industry," Dr. Cantor said.

Of course, this is only the first step in global drug research and development. Although this is a milestone, indicating that artificial intelligence has the potential to identify candidate drugs, it still needs years of clinical trials and millions of dollars of research before any potential drugs are approved for treatment.

AI technology can quickly identify effective DDR 1 kinase inhibitors. According to Insilicon Medicine

"This paper is an important milestone in our research and development of artificial intelligence-driven drugs. We have been engaged in AI synthetic chemistry since 20 15, but when Insilicon's theoretical paper was published on 20 16, everyone was skeptical. Now, this technology is becoming the mainstream, and we are glad that it is being verified in animal experiments. When these models are integrated into the whole drug research and development process, they are applicable to many target objects. We are working with leading biotechnology companies to further push the limits of synthetic chemistry and synthetic biology, "said Dr. Alex Zhavoronkov, the first author of the paper and founder and CEO of Yingshi Intelligent.

Editor Lu