News classification
Contact us
- Add: No. 9, North Fourth Ring Road, Haidian District, Beijing. It mainly includes face recognition, living detection, ID card recognition, bank card recognition, business card recognition, license plate recognition, OCR recognition, and intelligent recognition technology.
- Tel: 13146317170 廖经理
- Fax:
- Email: 398017534@qq.com
Is it reliable to use artificial intelligence to develop new drugs for cancer?
Is it reliable to use artificial intelligence to develop new drugs for cancer?
The National Cancer Institute estimates that 4/10 of Americans may be diagnosed with cancer at some point. 33% of cancer patients don't live longer than 5 years, so they have little time to find effective treatments. New anticancer drugs will take 10 years to reach the market, which involves animal testing, human trials and regulatory review - and the probability of success is only 7%. Is it surprising that there are fewer than 2000 drugs approved by the food and Drug Administration in the market? (not 2000 drugs for cancer treatment, 2000 drugs for all diseases)
Back in GANS, the first network has been trying to "cheat" the latter, allowing the latter to accept new molecules as legitimate drugs, and these two methods can better understand what cancer treatment should look like. Once they are tested with each other, these networks can be used to detect compounds to detect their anti-cancer potential. In this way, the Insilico Medicine research team screened 72 million chemicals from a public database. The choice of compounds in GANS have 60 to get cancer drug patent, which means that the network can accurately identify the drugs and other compounds, the choice is likely worth further study.
Compared with the in vitro (tube) experiment, the silico Laboratory (computer test) method is faster. The researchers did not find a new approach to cancer treatment, but the use of one million known compounds with therapeutic potential, and in just a month's time, the researchers can be reduced to 100 candidate list.
This method can not only promote drug development faster, but also bring more economic benefits to the research. Every drug that failed in the experiment would cost millions of dollars in labor and resources. A study by Journal of Health Economics estimated that the cost of each drug failed to increase by more than $1 billion 600 million per trial. With fewer, more accurate searches, researchers can save millions, even billions.
But not everyone has confidence in the application of silico testing. Mamoshina acknowledges that many cancer researchers who are using more traditional biological and chemical methods are not familiar with artificial intelligence, and the results of such experiments may not be good enough to trust them. "For them, this is a black box," she said. "It's really hard to understand, and that's why they're skeptical."
Like other cutting-edge technologies, hype may push the Insilico Medicine forward, but it also buys hidden dangers. Olexandr Isayev, an assistant professor at the University of North Carolina, focuses on developing artificial adjuvant drug discovery methods. He admits that for a technology that has not yet provided any substantial results, people may begin to be too excited. "Most of the published papers, including this paper, are purely computational," he said. "So, some predictions may be wrong. I really want to see that the discovery of the drug molecule "artificial intelligence discovery" has been successful for the first time."
The company has not licensed this technology in the form of software as a service model, but instead expanded its research into molecules that have been identified as potential anti-cancer agents on the internet. Once these compounds have been tested in vitro, they will be licensed to the pharmaceutical companies for further regulatory review, and if all goes well, marketing will be approved. In August, the company announced that Insilico Medicine was working with GSK, the pharmaceutical giant, to launch some new research technologies.
Insilico Medicine's belief in this new method is reflected in its decision to authorize the drug discovery, rather than the tool for discovering drugs. For the company, however, to prove that artificial intelligence really eliminates speculation that early detection of drugs is really ineffective, they will have to go back to the laboratory for testing.