Thousands of people, thousands of words of biomedicine | Interview with Dr. Xie Weidong, DM intelligence

November 20, 2020

Shortly after the outbreak of Xinguan in 2019, Guangzhou Zhirui Pharmaceutical Technology Co., Ltd. carried out new drug discovery work relying on (Zhong Nanshan team) of Guangzhou respiratory health research institute and School of life sciences of Sun Yat sen University. The research team plans to carry out the research and development of new drug resistant to 2019 nCoV through the "three-step" method of artificial intelligence screening, activity verification and drug evaluation. Using the artificial intelligence assisted virtual drug screening system, the high-resolution crystal structure of 2019 nCoV coronavirus main protease (3CLpro) was used to calculate and screen the listed drugs, and 40 potential effective compounds with better scores were obtained!

Recently, the Organizing Committee of BIO-PHARM2020 has the honor to invite Dr. Xie Weidong, chairman and general manager of DM intelligent Ltd., to participate in our special interview column of "thousand people, thousand words of biomedicine", to share some valuable opinions and experiences for us!

(Question 1)BIO-PHARM2020 Organizing Committee: at present, DM intelligence has developed several core pipelines. Please introduce the development progress of each model.

(Answer 1) Dr. Xie Weidong: DM intelligence focuses on the fields of tumor, nerve and anti infection, and has established 10 tumor target drug screening models, 2 neurodegenerative target drug screening models and 2 anti infective drug screening models. Among them, 5 pipelines developed rapidly, and lead compounds were obtained, which were specifically distributed in the fields of tumor (3 pipelines), neurodegeneration (1 pipeline), and anti infection (1 pipeline). As of August 2020, five target compounds with a total of 114 seedling compounds have been screened, including 40 compounds below μ m and 4 lead compounds with < 10 nm (including 2 compounds with < 1nm).

DM intelligence has cooperated with several well-known pharmaceutical companies. Among them, 3 targets have obtained their preclinical candidate compounds on their intended targets until they have signed an agreement to apply for ind approval, which is in the later stage of development.

Compared with homogeneous research, DM intelligence focuses on differentiated research and development, pays close attention to clinical needs and focuses on drug value. Globally, the progress of the above research and development pipelines is slow, and there are no drugs on the market, and most of them are concentrated in clinical phase I. The relevant transfer transaction also proves that the above pipelines are of great development value. With the smooth development of the pipeline, the market value of DM intelligence will also increase significantly.

(Question 2)BIO-PHARM2020 Organizing Committee: what are the advantages of artificial intelligence technology in the development of preclinical candidate compounds, such as traditional artificial screening and high-throughput screening? What are the challenges?

(Answer 2)Dr. Xie Weidong: the development of preclinical candidate compounds has the problems of long cycle and high cost. What's more, poor quality PCC is also an important reason for the high clinical failure rate. Nowadays, face recognition, automatic driving, speech recognition, semantic understanding have entered the homes of ordinary people, and artificial intelligence has been rapidly implemented in the fields of finance, public security, education, transportation, etc. It can be seen that algorithms and computing power have been developed rapidly. At the same time, the medical field has also changed from "big data" to "big data", which makes the application of AI in the medical field possible.

Compared with the traditional screening technology, AI technology can shorten the workload of three years to three months, exhaust more molecules, complete more dimensional data validation, and obtain more and better candidate molecules. However, while AI applications are blooming everywhere in other fields, the medical field often appears the scene that "those who understand medicine do not understand AI, and those who understand AI do not understand medicine". With the unity of knowledge breadth and depth, compound talents enterprises have great competitive advantages in the field of AI medicine.

(Question 3)BIO-PHARM2020 Organizing Committee: the pharmaceutical industry can be described as "one drug in ten years". The original intention of DM intelligence is to apply big data and algorithms to enable drug research and development, improve the accuracy of drug screening, shorten the research and development time, and reduce the cost of research and development. How does DM intelligence's drug discovery platform help customers increase the success rate of drug development?

(Answer 3)Dr. Xie Weidong: ensure the success rate of screening from many aspects. First, a deep understanding of drug research and development. For more than 30 years, we have been engaged in target research, compound research, preclinical research, clinical research, drug marketing and project management, and each link of the R & D process has been deeply involved. We believe that the drug discovery stage of innovative drugs is one of the main bottlenecks challenging the process of new drug development, and the data therein has great potential. The second is the effective combination of AI and drug R & D. At present, the company has established a number of large databases, and developed a set of its own algorithms. For specific targets, DM intelligence can deliver seedling compounds in only 3 months, with an average accuracy of 50%.

AI is only a tool. How to ask the right questions is the key. This is also the key difference between service-oriented AI new drug companies and R & D Cooperative AI new drug companies. Compared with the service-oriented AI drug company with computer as the core team, it is more necessary for the cooperative enterprises to put forward the correct questions. DM intelligence, as a R & D Cooperative AI new drug company, has its own unique vision on how to put forward the right questions, so it is more competitive in mining the potential value of new drugs.

DM intelligence can eliminate the bottleneck of new drug development, accelerate the process of new drug development, improve the efficiency of new drug development, shorten the circulation time of new drug to patients, and ultimately benefit patients and society.

(Question 4)BIO-PHARM2020 Organizing Committee: At the beginning of this year, boceprevir, an anti new crown compound, was screened by DM intelligence artificial intelligence new drug discovery platform. Later, it was confirmed by American scientists that its inhibitory effect was excellent. Can you introduce how DM intelligence is screened and validated?

(Answer 4)Dr. Xie Weidong: There is no time to delay the outbreak. DM intelligence research team carried out the research and development of new drug resistant to 2019 ncov through the "three-step" method of artificial intelligence screening, activity verification and drug evaluation.

DM intelligence calculated and screened 3clpro hydrolase protein with small molecules that have been on the market and under research. More than 50 antiviral candidates have been screened in one day. In February 1, 2020, novel coronavirus efficacy trials were published and purchased by 4 DM (intelligence, abacavir, Bicyclol Tablets and Da Lou Bernard Vee) jointly with the team of Boceprevir. On February 4, Academician Li Lanjuan announced that dalunavir can effectively inhibit coronavirus. The results of DM intelligence test showed that compared with the IC50 value of dalunavir, the inhibitory effect of boceprevir and bicyclol tablets screened from this platform was 2.7 and 0.93 times higher than that of dalunavir. A study published in the journal Cell Research by the school of medicine of South Florida University of health and the school of pharmacy of the University of Arizona in July also showed that boceprevir is the most suitable drug for inhibiting new coronavirus.

Through the DM intelligence artificial intelligence new drug creation platform, we obtained 9 validated and effective 2019 ncov 3clpro hydrolase inhibitors.

(Question 5)BIO-PHARM2020 Organizing Committee: Besides the domestic market, DM intelligence will also expand its global business. What kind of partners do you think the company needs to make further product / market layout?

(Answer 5)Dr. Xie Weidong: From the beginning of the company's establishment, wisdom medicine has been adhering to the belief of "urgent patients are in urgent need". From the perspective of drug value, smart Pharma is committed to the continuous and in-depth development of new drug pipelines, differential and accurate layout of drug racetracks.

DM intelligence makes full use of AI drug screening platform to improve R & D efficiency and quality, and provides technology integrated new drug R & D path. For enterprises with innovative drug demands and the same value concept as the company, they can make use of the difference advantages of both sides to carry out cooperation, maximize the potential and value of drugs, and accelerate the development and listing of drugs.

——The above source is BIO-PHARM2020