AI pharma is here

  New drug research and development or bid farewell to "a sword in ten years"

  Our reporter Sun Qiru

  AI face recognition technology can help catch fugitives from popular singers’ concerts, AI voice recognition allows everyone to use personal voice assistants, and AI medical imaging technology can help doctors complete image reading faster and better to find patient lesions... …When artificial intelligence technology has a long introduction cycle, high technical difficulty, and high human capital expenditure in the development of new drugs and other medical fields, the speed of human life-saving is gradually entering a new era.

  AI targets new drug development pain points

  Ten years of sharpening a sword is the key word in many innovative drug research and development stories.

Can a new drug be developed several years in advance at one-tenth of the cost?

Such a vision that seems to be a "Fantasy Night" is being driven into reality by artificial intelligence.

  Recently, British Silicon Intelligence, an artificial intelligence drug research and development company, announced that it has used artificial intelligence throughout the drug discovery process (including mechanism discovery, target discovery and finding new compounds), and successfully discovered a new mechanism drug, which is the first case in the world.

“It only took 18 months for research and development, and the research and development expenditure was about 2 million US dollars. Compared with the 10 years of traditional drug research and development, which cost tens of US dollars, AI-enabled drug research and development greatly shortened the drug development time and greatly reduced the cost.” Introduction by Ren Feng, Chief Scientific Officer of British Silicon Intelligence.

  “One of the most difficult steps and biggest mysteries in drug discovery lies in the verification of targets, especially the identification of targets that have a strong influence in the clinical environment. Through the efforts of artificial intelligence, Anglo Silicon has successfully solved the biggest problem in drug discovery. One of the mysteries." A senior practitioner with more than 20 years of experience in the field of drug research and development commented.

  What is a target?

What does it have to do with new drugs?

In popular understanding, if the disease is compared to a lock, then the target is the lock cylinder. If the lock cylinder is found and the three-dimensional structure of the lock cylinder is researched, people can further equip it with a proprietary lock according to the three-dimensional structure of the lock cylinder. The key, the new medicine is the exclusive key.

  Introduced by Anglo Smart R&D personnel, the team has spent several years building and integrating hundreds of artificial intelligence models. Each model is responsible for a specific task and integrated on a platform that can generate hypotheses and choices. Targets, generating compounds and predicting the results of clinical trials, so as to greatly improve the efficiency of research and development in the corresponding links.

  "Three highs" field leads to the end of giants

  Tirelessness, stable accuracy, super learning ability, unaffected by emotions and other factors... With the gradual implementation of artificial intelligence technology in areas such as autonomous driving, medical imaging, and public safety, the many advantages of AI are gradually becoming familiar to people.

  In areas related to human life and health, AI has also begun to play its role.

Previously, in the field of AI medical imaging, artificial intelligence has been able to help humans find suspicious lesions in massive medical images.

  New drug research and development is one of the most risky, complicated, and time-consuming technical research fields in human development.

Generally speaking, the development of a new drug needs to go through the stages of drug target determination, lead compound screening, lead compound optimization, and clinical trials. During this period, it not only requires many experts in the fields of biology, chemistry, and medicine, but also bears a huge risk of failure.

According to statistics, it takes an average of US$2.6 billion for pharmaceutical companies to develop a new drug, and the development time of up to 10 years, the failure rate of drug development is usually more than 90%.

  Large capital requirements, long R&D time, and high failure rate-this "three highs" field, which is extremely important to human life and health, urgently needs the help of artificial intelligence to improve R&D efficiency, reduce costs and shorten the cycle.

In fact, the application of artificial intelligence technology has been proven to be helpful in almost every step of the drug discovery process, especially in the stage of disease hypothesis and target identification.

  At present, there are more than 200 companies around the world that use AI to assist drug development.

Since the second half of last year, AI drug research and development has ushered in an investment-intensive period.

At the same time, technology giants such as Baidu, Tencent, Huawei, ByteDance, etc., in the field of non-medical origin, have also deployed AI medicine.

  At the beginning of this year, Baitu Shengke, led by Baidu founder Robin Li, launched the "One Million Leadership Program" and "One Million Youth Leadership Program", which will use annual salary of 1 million US dollars, 1 million yuan annual salary and other technical platform support to attract Biotechnology + AI technology cross-border integration of talents.

Li Yanhong predicts that in the future, artificial intelligence has great potential in public health monitoring, new drug research and development, and disease diagnosis.

  Prior to this, Bytedance and Huawei both began to recruit biomedical-related positions, and Tencent released the first AI-driven drug discovery platform Yunshen Zhiyao independently developed by it.

  "AI+Medical" has just started

  "From the gradual derivation of AI + imaging to drug development, AI + medical is currently only taking the first step from 1 to 10, and the potential for the next 20 years is huge. The combination of AI and humans may promote the medical industry more than humans in the past few thousand years. Kai-Fu Lee, chairman and CEO of Innovation Workshop, said that when AI medical is scored from "1 to 10", only 1 point is currently available.

  Kaifu Li said that the future scenarios of AI + life sciences are worth looking forward to. At present, the proportion of robots participating in surgery is close to 20%, and some of them are dominated by robots.

For example, he said that after the emergence of nano-robots in the future, they can enter the human body to help humans fight cancer.

  According to Liang Yongyu, the managing partner of Qiming Venture Capital, AI-enabled drug research and development can not only shorten the time of drug research and development, but also greatly reduce the cost.

In the future, drug development is one of the most important and largest application scenarios of artificial intelligence, and artificial intelligence is one of the most significant technological dividends of drug development.