AI tracks the evolution of the global coronavirus, big data to create precise pandemic prevention, the key report of the "All-round AI x Pandemic Prevention Online Forum"

Posted by Gene Online on May 07, 2020

In view of the continuing global COVID-19 epidemic and the possibility of a long-term anti-Japanese war, the Ministry of Science and Technology subsidizes the full-scale health care sub-center, the National Taiwan University Artificial Intelligence and Robotics Research Center, the Taiwan University Student Medical Ethics Center, the Taiwan Development Research Institute, and the consortium. "TAIWAN is Helping: All-round AI x Epidemic Prevention Online Forum" co-organized by Ying Rui Biomedical Consulting Co., Ltd., Taiwan Suicide Prevention and Treatment Association, American Tuce Intelligent Technology Co., Ltd., and Taiwan Medical Management Association was held at the National Taiwan University Jisi Conference Center. The forum extended the four major aspects of epidemic prevention topics, including: virus and AI application analysis, detection and treatment development, public health and psychological trauma, and legal and medical ethics. 16 domestic and foreign industry, government, and academic related institutions were invited, including 16 well-known domestic and foreign institutions. Scholars and experts serve as speakers, sharing their research findings. In addition to attending the site to listen, this forum also has an online live broadcast. It is hoped that more people can participate in the forum in response to the reduction of contact measures caused by the epidemic, and I hope to contribute to the epidemic prevention of Taiwan and the world.

First of all, Xu Youjin, Undersecretary of the Ministry of Science and Technology, mentioned in a speech that in the past experience in combating SARS, Taiwan's science and technology industry has greatly assisted. AI and big data technologies are beneficial to human beings. In this wave of epidemics, they are helpful for epidemic prevention, drug selection, and vaccine research and development, and can track viral infections. AI smart medical care is the future trend, and it can help the epidemic through cross-field and cross-international medical treatment.

Ching-Yung Lin, CEO: The global COVID-19 virus can be divided into eight categories, and AI can monitor and track the evolution of the virus

Virus and AI application analysis play an important role in both epidemic monitoring and technology research and development. The organizer invited Dr. Ching-Yung Lin , CEO of Graphen, Inc., with the topic of "Graphen COVID-19 Virus Evolution Monitoring" , To interpret the genetic sequences of more than 12,000 strains of viruses around the world and divide them into eight categories. As of April 29, 2020, 12,294 virus strains have been sequenced, including 6205 variants. Dr. Lin said that since the end of 2019, several clusters of virus strains have been caught growing and sequenced for them. Because of the theory of evolution and the survival of the fittest, the virus will begin to mutate. And the more infected people, the more mutation types; the longer the time, the more mutation points; on average, the new coronavirus adds a stable survival mutation point every week. Dr. Lin explained that the genome is like a part of a machine. It is a piece of software as well as a piece of hardware. Genetic changes will directly change the protein. The COVID-19 genome produces 26 proteins, and a protein is like a part. When the parts change their tone and produce different functions, they change more and more and produce a variety of new coronaviruses. The reason why the COVID-19 virus is so strong is its replication ability and camouflage ability, which makes the body's immune system think that it is a foreign object, which has been spreading during the incubation period, and it is easy to explode after replication. Alt Text Online remote connection speech by Dr. Ching-Yung Lin , CEO of Graphen, Inc.

Dr. Lin further analyzed the eight major COVID-19 viruses that attack the world, which are divided into eight categories: A (divided into A1, A2), B, C, D, E, F, G, and H. Type A is mainly distributed in China, and genetic analysis shows that all viruses are likely to have evolved from these two strains (A1) and (A2). One of the strains (A2) is almost the closest to the virus in bats and pangolins. The main distribution area of Class B is also China, all of which are all kinds of virus strains directly evolved from A1. These changes affect various proteins, but these mutations have not been spread on a large scale in the existing sequences. The C type is mainly distributed in Europe and is characterized by spike protein (S protein) mutations. The S protein is the key to the virus infecting humans. However, the mutation site is not on the RBD that binds to human ACE2. Class D is mainly distributed in England, Netherlands, and Hong Kong, and is characterized by NSP2, NSP3 and ORF3 protein mutations. NSP3 is responsible for cutting apart the various proteins of the virus, and is responsible for the proteins that affect the infected cells. The main distribution areas of the E category are the western United States and Canada, which are characterized by ORF8 variants and most of them also have two NSP13 variants. The function of ORF8 protein is yet to be studied. NSP13 may be a virus used to unfold RNA. The main distribution areas of Class F are West, Australia, South Korea, and China, and are characterized by various virus strains directly evolved from A2. These changes affect various proteins, but these mutations have not been spread on a large scale in the existing sequences. The main distribution areas of the G category are Europe and South America, and are characterized by the variation of three consecutive sites on the S protein and its nuclear protein (N protein). The N protein is used to protect the shell of the viral gene RNA. The main distribution areas of the H category are France and the East of the United States, and are characterized by S protein mutations and another ORF3 (de-jumping expert) protein mutation. The function of ORF3 is to puncture the host cell membrane, allowing the virus that replicates in it to spread. Alt Text Classification tree diagram of the eight major COVID-19 viruses in the world Alt Text Distribution regions and characteristics of the eight major types of COVID-19 viruses in the world

Dr. Lin also stated that his company Graphen is currently monitoring four kinds of mutations, including Any Mutation, Missense Mutations, Silent Mutations, and Radical Mutations, and any range is possible. monitor. In addition, due to the COVID-19 pandemic, the global AI medical research and development will reach more than 2 billion U.S. dollars by 2025. Finally, I hope to solve human problems through advanced AI technology.

Extended reading: Highlights of the Taiwanese Epidemic Prevention Diplomacy Forum: How can medical, public health, and AI defeat Wuhan pneumonia?

Deputy Director Zhuang Yaoyu: Taiwan needs to strengthen the establishment of a new crown case gene sequencing database. The host's gene sequencing also needs to be paid attention to

Genome research expert and Deputy Director of the Institute of Biomedical Sciences (ITRI) Zhuang Yaoyu shared RNA sequencing and data analysis, gene database and analysis tools, and COVID-19 sequencing with the topic of "Pandemic and Gene Big Data Application" Database, virus and host genetics, AI technology of bioinformatics, ITRI's COVID-19 diagnosis system, etc. Deputy Director Zhuang pointed out that the UK has already completed the genetic sequencing of more than 10,567 new coronavirus strains. Taiwan currently has 69 in the GISAID database, which is less than one-sixth of the number of confirmed cases in Taiwan. Taiwan needs to strengthen its efforts in this regard. Some efforts, and in line with international standards. He emphasized that in addition to the genetic sequencing of the COVID-19 virus itself, the "host genetics" (host genetics) is also very important, that is, the genetic sequences of the infected patients themselves. More research in this area can be done in the future. The ACE2 receptors on Taiwanese may vary from person to person, so you should pay more attention when doing research. He concluded that because diseases are very complex, multi-omics is necessary to provide a comprehensive interpretation, rather than just using sequencing. Domain knowledge is necessary for big data analysis and AI. Studying host genetic variation is as important as the COVID-19 virus sequence. ITRI's BDL has developed two nucleic acid-based COVID-19 diagnostic systems. Alt Text Zhuang Yaoyu, Deputy Director, Institute of Biomedicine, Industrial Technology Research Institute

Professor Li Youzhuan: Telemedicine can help prevent epidemics

Li Youzhuan, Distinguished Professor of the Institute of Medical Information, Taipei Medical University, who will be the chairman of the International Medical Information Association in 2021, sharing, prediction (Predict), telemedicine (Telemedicine), big data sharing (Sharing / Big Data), information transparency (Transparency) is the four key to epidemic prevention. Now is a good time for the world to learn about the value of shared medical big data, and AI can help. From the perspective of "predict", AI can be used for precise epidemic prevention, because psychological fear is more terrible than the new crown virus, and the reason for panic is that the virus occurs randomly, regardless of gender, age, occupation, nationality and race; asymptomatic It can be transmitted for a long time (contact/droplet/aerosol) to the recovered and recovered; it can be attached for a long time and "live" on the surface of inorganic bodies. AI can screen high-risk individuals in the early stage. As for who should be quarantined, we should focus on precision and individualization. Do. Furthermore, the strongest line of defense is that as long as the medical system does not collapse, large-scale community infections will not easily occur. Provide the medical system with the strongest resources, best protection, and sufficient manpower, so that AI can break fears, perform pre-hospital screening, and have a systematic digital rapid screening, which can reduce medical collapse. AI can also do early detection, predictive deployment, and telemedicine. Regarding telemedicine, although Taiwan has related communication diagnosis therapies, it still has shortcomings, but at least Taiwan can do pre-screen consultation for the epidemic situation. If there are fewer outpatient clinics in Taiwan, costs can be reduced. Alt Text Li Youzhuan Distinguished Professor, Institute of Medical Information, Taipei Medical University

Vice President Chen Xiuxi: Taiwan needs preventive vaccines and may consider implementing a serological monitoring program

Deputy Dean Chen Xiuxi of the National Taiwan University School of Public Health gave a speech on "The Present and Future of Taiwan's Epidemic Prevention Deployment." COVID-19 is a global pandemic. There are not many confirmed cases in Taiwan and the fatality rate is also very low. Stop (border management and quarantine) and harm reduction plans (including reduction of large gatherings, personal protection such as wearing masks and frequent hand washing, environmental protection, etc.). However, as the epidemic slows, countries gradually unblock them, and Taiwan's epidemic prevention is a challenge. challenge. For example, Chen Xiuxi, the Cyber-Condolence digital grave sweeping during the Ching Ming Festival was not bad. He also mentioned that Australia has done a good job in epidemic prevention. They have launched a social distancing mobile app COVIDSAFE, which may be a reference for Taiwan. In addition, he also suggested that possible IoT applications can be used in COVID-19 quarantine and home medical care. Finally, regarding the future challenges of Taiwan and the world, Chen Xiuxi pointed out that Taiwan can infect many hosts and needs preventive vaccines to achieve herd immunity. It also needs therapeutic vaccines and antiviral drugs to reduce infection and accelerate recovery. In order to maintain immunity, a Serological Test Surveillance (Serological Test Surveillance) can be implemented. Alt Text Chen Xiuxi, Deputy Dean of the School of Public Health, National Taiwan University

Researcher He Meixiang: Neutralizing antibodies are an indicator of immunity and can be used for treatment. It is recommended to develop fast screening agents that can test neutralizing antibodies.

Researcher He Meixiang from the Institute of Biomedical Sciences, Academia Sinica, an expert who has participated in anti-SARS operations, shared the strategy of offensive and defensive epidemics with the topic of "Strategies for Epidemic Attack and Defense". What about the next wave of epidemics? First, she proposed to answer several important scientific questions first, including whether there is immunity after infection, whether viruses and antibodies can coexist, and when COVID-19 patients are or are not infectious. In the meeting, she answered these three questions separately. First, the infected patient has immunity. Second, viruses and antibodies can exist at the same time. Third, the infectivity of COVID-19 patients reached a peak on the day of onset, and then gradually declined, and after the emergence of antibodies, the infectivity has been greatly reduced, which indirectly explains why there is a virus (nucleic acid positive) in the rock ship cluster outbreak. ), but not infectious.

He Meixiang further proposed that neutralizing antibodies are an indicator of immunity and are also protective. Patients diagnosed with COVID-19 will produce neutralizing antibodies under normal circumstances, but there are still no neutralizing antibody reagents for general screening. She suggested that Taiwan should develop A quick screening agent that can test neutralizing antibodies. In application, neutralizing antibodies can be used for treatment, and can be used as an indicator of entry and exit exemption from quarantine. In addition, a small number of data show that asymptomatic infections may not produce antibodies, or antibody production is slower, and how to deal with asymptomatic infections needs to be discussed separately (now enter the negative pressure isolation ward).

He Meixiang also said that if herd immunity is achieved, the epidemic will decline. There are two methods, one is natural infection, and the other is vaccinations. Regarding vaccines, as soon as the third quarter of 2020, "emergency use" will be used in small quantities for medical staff; regarding vaccines for universal use, due to the applicability of emergency medicine certificates for universal use or production capacity, the time course of universal use is difficult to estimate , It is super optimistic that the vaccine for universal use will not be available until the third quarter of 2021. Alt Text Researcher He Meixiang, Institute of Biomedical Sciences, Academia Sinica

The forum also invited Dr. Yixiang Xu, Associate Professor of Harvard Medical School, who is the director of the Marcus Institute for Aging Research Genetic Epi Program (Genetic Epi Program), and Dr. Limin Huang, Director of the Taiwan Infectious Diseases Association and Director of the Pediatric Department of National Taiwan University Hospital, and Ying Yunsheng Zhu Fenyu, President of Medical Consulting Co., Ltd., Wang Yutai, team leader who participated in the COVID-19 Global Epidemic Map'' project of the State Grid Center of the National Research Institute, and the Precision Medical Diagnostic Guidelines Technology Group of the Institute of Biomedical Research, participated in the development of theHandheld Nucleic Acid Molecular Quick Screening System'' Team Leader Chen Tingshuo, Director Wang Zhihong of the Center for Policy and Achievement and Prevention of Stanford University who wrote an article in the authoritative medical journal "Journal of the American Medical Association" (JAMA) in March this year to share Taiwan's epidemic prevention practices, and Director Wang Zhihong of the National Suicide Prevention Center, National Taiwan University Associate Professor Wu Jiayi of the Department of Nursing, Professor Cai Fuchang, Chairman of the Taiwan Society for Clinical Research Ethics Review, Researcher He Zhixing, Academia Sinica Research Institute of European and American Research who is a member of the Taiwan Human Biodata Database Ethics Committee, and founder of "AI Law Review Network" Li Chongxu, director of the Institute of Medical and Biotechnology Law of Taipei Medical University, joined the forum to provide professional opinions.

Director Fu Licheng: AI and big data can help prevent epidemics

At the end of the forum, Director Fu Licheng of the Ministry of Science and Technology subsidized the full-scale health care sub-center concluded that the COVID-19 epidemic has caused this wave of anti-epidemic waves at home and abroad. The Chinese government has SARS experience so that it can respond as soon as possible, and it has world-class medical care. The public health and the people have a high degree of epidemic prevention concepts, which has made Taiwan's achievements in this epidemic prevention world attention. This time, the Ministry of Science and Technology subsidized Taiwan's full-scale health care sub-center, and invited opinion leaders from all walks of life, across units, fields, and at home and abroad to participate in the forum, hoping to exert its influence.

Director Fu further explained that looking at COVID-19 from the perspective of AI and big data analysis, whether it is virus genes, medical treatment of cases, or public health and information, the monitoring of individual and group epidemics and the identification of high-risk groups are all relevant The benefits also enable us to further accurate screening and treatment; today, some speakers mentioned that general screening will waste too much national financial resources, so accurate screening should make good use of big data. In addition, my country's epidemic prevention results have resulted in only more than 400 COVID-19 cases in the country, but from the perspective of big data, this data is slightly less, so we should actively cooperate with countries around the world, especially the current confirmed cases account for two-thirds of the world. The European and American countries have given Taiwan the energy to participate in international research, and also assisted in expanding participation in the development of the epidemic prevention clinical industry and opening up the international market.

In addition, Director Fu also mentioned that this forum discussed the aspects of human and social sciences such as legal ethics, social impact and mass psychology derived from the epidemic prevention and public health policy, which also need to be paid attention to, such as epidemic risk assessment and epidemic models. Medical ethics related to prediction and epidemic prevention all need to be taken seriously. Although we are talking about AI technology today, all aspects of the issue need to be paid attention to in order for the national society to have a good society and life under the attack of the epidemic. It is hoped that today's discussion will help the government formulate future public policy on epidemic prevention and assist the people in facing the impact of epidemic prevention, which will further influence the international community's imitation of Taiwan's epidemic prevention.

Finally, Director Fu said that although the domestic epidemic has eased, the international epidemic is still very severe. It is necessary to prepare for a long-term war and respond to other epidemics that may occur in the future; the Ministry of Science and Technology subsidizes the full-scale health care sub-center to also plan and prevent epidemics. It is hoped that the theme of the forum can cooperate with the world to achieve all-round goals.


Originally in Chinese