cell analysis | Molecular map | IND
biometrics | Gene sequencing | AlphaFold
Under the background of rapid development of cell analysis, visual recognition, biometrics, gene sequencing, IND, and AlphaFold, various emerging technologies that benefit mankind have begun to appear. Recently, the 20th anniversary theme summit of the "Top Ten Breakthrough Technologies" of "MIT Technology Review" was successfully held in the Future Science and Technology City, Yuhang District, Hangzhou.
Academician of the Chinese Academy of Sciences, Chairman of the Development Committee of Zhejiang University, Professor Yang Wei of Zhejiang University, Academician of the Chinese Academy of Engineering, , Professor Academician of the Chinese Academy of Engineering, , Professor Professor of the School of Environment, Tsinghua University, Academician of the Chinese Academy of Engineering, Dean of the Institute of Intelligent Industry of Tsinghua University, Zhang Yaqin and other nearly 100 top domestic and foreign scientists, industry leaders, and business elites were invited to attend.
The breakthrough technologies released this time include: COVID-19 oral drugs, practical fusion reactors, termination codes, AI protein folding, PoS permission proof, long-term grid energy storage batteries, AI data generation, malaria vaccines, carbon removal factories and COVID-19 mutation tracking.
new crown oral medicine
Since the outbreak of the new crown epidemic, major pharmaceutical companies and scientific research institutions around the world have been committed to developing vaccines that can effectively prevent new crown virus infection and effective therapeutic drugs for new crown .
According to statistics, as of early May, there were more than 1,200 new crown drugs under development worldwide, and nearly 50% of the projects in the IND and above R&D stage accounted for nearly 50%. 15 applications were submitted for listing, involving more than 1,000 companies; at present, more than 50 drugs (including vaccines) have been approved for the indications of new crowns, including 12 small molecule chemical drugs and more than 30 biological drugs.
From the perspective of mechanism of action, among the approved small molecule chemical drugs, there are three RdRp inhibitors (RNA polymerase inhibitors since RNA), namely Gilead remdesivir , Toyama Chemical's fapiravir , and Merck's molnupiravir; in addition, it also includes Pfizer 3CL protease inhibitor Paxlovid, COVID19 replicase polyprotein 1a inhibitor combination drug Nematvir + ritonavir , Incyte's JAK inhibitor baritinib, etc. Among the major biopharmaceutical categories, the number of drugs with a mechanism of action of COVID19 spike glycoprotein regulators accounted for the highest proportion (about 50%); in terms of therapy types, in addition to vaccines, neutralizing antibodies accounted for the majority, including sotrovimab, cassirimab + Idzumab, barnivimab, etc. In addition, nucleic acid drugs such as tozinameran, elasomeran, ZyCoV-D are also approved for the treatment of the new crown.
At present, there are two oral small molecule specific drugs for the new crown that have been launched in the world, Molnupiravir of Merck and Pfizer's Paxlovid. There are more than 10 new crown drugs (including vaccines) on the market in China, including oral drugs fapiravir, nematodevir + ritonavir, neutralizing antibodies ambavimumab + romisvirumab, etc., most of which are emergency use authorization/conditional approval.
Compared with remdesivir and neutralizing antibody drugs that require injection, oral small molecule drugs have more advantages, and therefore has become a popular track in the development of new crown drugs in the world.
Advantages of oral small molecule special drugs for COVID-19:
1. The patient has high tolerance and good compliance, which is convenient for inhibiting the proliferation of the virus in the early stage of infection and avoiding the conversion to severe
2. The price is low. Mumpinavir is priced at US$700 per person in the United States, which is only one-third of the neutralizing antibody
3. It is easy to transport and easy to distribute. Compared with antibody drugs that require intravenous injection, oral small molecule drugs are undoubtedly much more convenient. In underdeveloped countries with severe epidemics and backward medical conditions, oral small molecule antiviral drugs are more practical
From the perspective of domestic corporate competition, there are currently more than 100 companies in China participating in the research and development of new crown drugs, involving more than 150 R&D projects. The domestically produced oral drugs for new crown under development include more than 10 models such as Azvdine, VV116, Purkluamide, SIM0417, and RAY003. The listed companies involved include Junshi Biologics , Kaituo Pharmaceutical, Xiansheng Pharmaceutical , Zhongsheng Pharmaceutical , etc.
Judging from the progress of drug research and development, there are 6 in the clinical trial stage, among which the three fastest-progressed are Azvdin of Real Bio, VV116 of Junshi Biologics, and Purkluamide of Kaituo Pharmaceutical. The first domestic oral small molecule special drug for the new crown is basically locked in these three drugs.
1. The development of "excessive speed" VV116
VV116 is a new oral nucleoside anti-SARS-CoV-2 drug. It is an RdRp inhibitor that can inhibit SARS-CoV-2 replication. Currently, it has obtained emergency use authorization in Uzbekistan . This is another new crown oral drug approved for marketing in the world after Merck and Pfizer's new crown oral drug. It was jointly developed by Shanghai Institute of Materia Medica, Chinese Academy of Sciences, , Wuhan Institute of Virology, Chinese Academy of Sciences, Xinjiang Institute of Physics and Chemical Technology, Chinese Academy of Sciences, , Wangshan Wangshui Biopharma Co., Ltd., and Central Asia Drug Research and Development Center of Chinese Academy of Sciences.
According to Junshi Biologics, VV116 achieved the primary and secondary efficacy endpoint preset of the protocol in a phase III registration clinical study (NCT05341609) comparing Nematvir tablets/ritonavir tablets (i.e. Paxlovid) for mild to moderate novel coronavirus pneumonia with progression to severe mortality and high-risk patients with severe mortality. It also issued an announcement stating that it will communicate with the drug regulatory authorities to submit applications for new drug marketing in the near future.
After its research results were released, industry discussions and doubts about the clinical trial of VV116 also flocked to it. The reason is that the research and development process of its clinical trials is too fast. The industry believes that its research directly shifts to Phase III clinical practice in a strict sense, and compared with the research and development of VV116, the time for development of VV116 is relatively short. In addition to the "speeding" of the R&D process, VV116 has also caused some controversy in terms of safety and patents.
Although there is some controversy, Junshi Biologics said that the research has reached the main endpoint preset of the clinical plan. At least from the point of view of the time of continuous clinical recovery, the efficacy of VV116 is no less than PAXLOVID. In general, its successively published experimental data boosted the market's confidence in the subsequent development of VV116.
2. Azfuding, which is hot without approval,
Azfuding was originally a drug for treating AIDS . In July 2021, it was approved for marketing for use in combination with nucleoside reverse transcriptase inhibitors and non-nucleoside reverse transcriptase inhibitors to treat adult HIV-1 infected patients with high viral load, becoming the first anti-HIV drug in my country that truly possesses independent intellectual property rights.
After the outbreak of the new crown epidemic, real biology carried out research on Azvdin's treatment of new crown pneumonia. From the perspective of targets of action, azvdine acts on RdRp (RNA polymerase), which is different from Pfizer's Paxlovid acts on 3CLpro (3C-like protein), but is similar to Merck's Molnupiravir.
In mid-April, Real Biology released some data on Azvding to fight the new coronavirus. Judging from the results of phase II clinical and some phase III clinical trials, the time for turning azvdine nucleic acid into negative nucleic acid is 3-4 days, the average time for medication is 6-7 days, and the average time for discharge is 9 days. Severe and mild symptoms are similar in treatment, and are also effective in patients who are ineffective in other drugs, and unlike Paxlovid, it is not necessary to take it in the early stages of infection with the new coronavirus.
On May 12, Junshi Biologics disclosed its price to the public for the first time. According to media reports, the price of VV116 in Uzbekistan is US$185, which is approximately RMB 1,243. As the results of the anti-COVID-19 clinical trials have not yet been released, Real Bio has finalized three manufacturers and distributors for Azfuding (China Resources Double Crane, Xinhua Pharmaceutical , and Aoxiang Pharmaceutical), and is therefore jokingly called "one woman and three marriages" by the industry.
3. The twists and turns of the prokrutamine
pcortamine was originally the second generation of AR antagonist used by Kaituo Pharmaceutical to treat prostate cancer . After the outbreak of the new crown epidemic, clinical trials confirmed that the drug has a therapeutic effect on the new crown. In early 2021, in the Phase III clinical trial in Brazil, the risk of death of severe COVID-19 patients could be reduced by 92%, and was once regarded as the "people's hope." Later, it caused considerable controversy because the interim analysis of the Phase III clinical trial did not achieve statistical significance.
In April 2022, Kaituo Pharmaceutical released key data on the third phase of clinical trial, specifically pointing out that "prikrutamide effectively reduces the hospitalization/mortality rate of new crown patients, especially for all patients who have taken the medicine for more than 7 days, as well as middle- and elderly patients with high risk factors, reaching a 100% protection rate, which is statistically significant." From clinical failure to 100% effectiveness, it is worth looking forward to whether prokrutamide can turn the tables in the competition for heroes.
In addition, domestic new crown drugs under development are still in the early stages of research and development: FB2001 of Frontier Biotechnology-U (688221.SH), SIM0417 of Xiansheng Pharmaceutical (02096.HK), ASC10 and ASC11 of Geli Pharmaceutical-B (01672.HK), 3CL protease inhibitor of Guangsheng Zhonglin/ WuXi AppTec (603259.SH), RAY003 of Zhongsheng Ruichuang, etc.
Utility Fusion Reactor
Commonwealth Fusion Systems researchers slowly charge a 10-ton D-magnet and increase the field strength until it exceeds 20 Tesla (T). This is a new record for similar magnets. The company's founder said the feat solved the major engineering challenges faced in the development of a compact, inexpensive fusion reactor.
For decades, nuclear fusion power generation has been a dream of physicists. At temperatures far above 100 million degrees Celsius, just like in the sun, nuclei fuses together, releasing a lot of energy in the process. If researchers can achieve these reactions in a controlled and continuous way on Earth, it can leverage almost unlimited fuel sources to provide cheap, sustained, carbon-free sources of electricity.
In one of the methods, magnets are used to confine ions and electron gases, so-called plasma , to a donut-shaped reactor. A stronger magnet means less heat loss, allowing more nuclear fusion reactions to occur in a smaller, cheaper facility. This change is more than a little bit: magnetic field strength doubles, and the volume of plasma required to produce the same energy is reduced by 16 times.
Although research over the past decades has cost billions of dollars, no one has built a nuclear fusion plant that produces more energy than the reactor . However, Commonwealth Fusion Systems and its supporters are full of hope, and other fusion startups and research efforts have also reported recent progress.
Commonwealth Fusion Systems is building a factory to produce magnets on a large scale and lay the foundation for prototype reactors. If everything works, the startup plans to provide fusion energy to the power grid in the early 1930s.
Termination Password
Single long-term use password will bring crisis to your own property, which is also the reason why security researchers encourage everyone to change their passwords frequently and try to set different passwords for different devices. However, some people may say that if you set too much password, it is easy to forget it. So what should I do? It is in this case that biometric technology came into being.
Although biometric technology is very useful for improving security, its transformative benefits go beyond that. Without the need to remember cumbersome passwords, biometrics can significantly improve the customer experience. As businesses begin to recognize the security risks and poor user experiences of knowledge-based authentication, multimodal biometrics and verification will be applied.
Despite this, things are not that simple. As early as 2014, German hacker Jan "Starbug" Krissler demonstrated how fingerprints were forged through high-resolution photos of different people's hands, and emphasized the potential vulnerabilities of the technology.
, coinciding with the release of Apple's iPhone 5s, Stargbug immediately became a hot search because he successfully "deceived" Apple's TouchID sensor and unlocked the phone. It is understood that the fingerprint was extracted through the stains on the screen, and the phone was unlocked.
is not just fingerprint recognition, but actually voice recognition also has certain risks.For example, criminals or hackers will consciously record the victim's voice and then use this to avoid the control of authentication. Perhaps, soon we will see a person's face being "reverse-engineered" (an algorithm that only requires some 2D images to complete), and with the help of a 3D printer, the scammers can walk around with the victim's mask and even take thousands of dollars from the ATM.
In the author's opinion, not all defects of biometric technology are discovered by external sources. Similar to other technologies, as more and more technology is adopted, the essential shortcomings will become clearer. For example, despite improvements in matching accuracy, false positives still plague their implementation, which is difficult to avoid in the development, configuration and deployment technology process.
But it is also this misjudgment that has occurred in the police circle that urges us to stop and think. Perhaps the biggest drawback of this technology is that the details of biometrics are static. If the password is stolen, there is still a chance of modification, but when a database full of information is destroyed, the individual's fingerprint, iris or other facial features cannot be replaced.
Therefore, while biometrics are an exciting new technology, its use must be implemented in a calm, planned and strategic way. For example, when verifying identity, biometric technology must use at least two of them; when helping 3 police officers to track down criminal suspects, the results must be confirmed by human analysts, and, more importantly, all biometric data must be stored.
AI protein folding
0 A few days ago, AlphaFold, a big star in the calculation biological world, made another major breakthrough. It has been able to predict the structure of 214 million proteins in more than one million species, covering almost all known proteins on Earth. The emergence of AI has greatly changed the pattern and efficiency of protein prediction. At present, all universities and enterprises have relevant layouts, and relevant startups in my country have shown explosive growth from 2017 to 2021, and most of them have received high- financing .
Not long ago, Internet giant Meta updated the protein model ESMFold. It can predict complete protein structure directly from single-sequence language model representations, with accuracy comparable to AlphaFold and reasoning speeds that are an order of magnitude faster. Huashen Intelligent Pharmaceutical, a domestic AI innovative drug company led by Peng Jian, has also achieved the latest breakthrough: OmegaFold uses a single sequence to solve the 3D structure of the protein. Even if the protein is artificially designed, its functions can be determined through AI prediction of the 3D structure.
1. Domestic computational biology showed explosive growth from 2017 to 2021
computational biology, essentially speaking, it is to solve the biology problem through computational means. Specifically, it is to build algorithms and models based on different types of biological data (such as concentrations, sequences, images, etc.), so as to understand the biological system itself (such as molecules, cells, tissues, and organs, etc.), and to promote related research and application disciplines.
is divided from application. Currently, the main areas of implementation include sequence analysis, structural and functional analysis, biomolecular dynamics, system modeling, evolution and population genomics, correlation networks...
Take AlphaFold2 as an example. It predicts protein structure based on gene sequences and belongs to the category of structural and functional analysis.
You can see that computational biology belongs to a subject of tool nature. To some extent, this determines that there are no computing biology companies on the market in a strict sense, but appear in the name of AI pharmaceuticals, omics and , precision medicine, etc. This is particularly obvious in our country.
At present, China takes AI pharmaceuticals as its core scenario. Not only university institutions (School of Life Sciences, Peking University Frontier Interdisciplinary Research Institute, etc.), but also major Internet manufacturers (Ali, Baidu, , Huawei , etc.) have related research and layout. Related startups showed explosive growth from 2017 to 2021, and have all received high financing. This situation is also reflected in foreign countries.
According to the report "Investment and Exit Trends in the Medical and Health Industry" of Pufa Silicon Valley Bank, the amount invested in computing biology companies in 2021 reached US$5.9 billion (i.e. 39.7 billion yuan) increased by as much as three times a year, exceeding twice the investment of non-computing biology companies.
From the perspective of business model, the entire industry is dominated by 2B, mainly algorithm authorization, biological assets and software use. my country mainly includes the first two, but given that software platforms and pioneer projects can form a closed loop of technology and business iteration.
After a large number of advantages in self-developed algorithms appear, the proportion of software platforms will increase significantly. Foreign countries have begun to use its computing biology services to foreign countries through packaging subscriptions and billing according to usage.
2. The emergence of AI or deep learning has brought a turning point to computing biology
The reason for the popularity of this track today is: firstly related to the explosive growth of deep learning in recent years; secondly, the recently emerging concept of AI for Science, which has made the symbol of AI's implementation in the field of biology, computing biology, a trend. The huge potential brought by the combination of AI and traditional scientific research is expected to bring a new scientific revolution ; finally, for biology itself, traditional experimental and analytical methods are difficult to fully develop massive biological data. It is indeed necessary to solve the problem by comprehensive tools such as interdisciplinary science, which take into account multiple subdivisions.
So, what value can computational biology bring to biology? The "In-depth Industry Report on Computational Biology" believes that it is divided into two major areas: scientific research and application.
The most direct role of computational biology in scientific research is to replace and even surpass experiments. Compared with traditional biological experiments with limited accuracy such as operational level, experimental equipment, observation level, etc., computer-based computational biology is not only cheaper and faster, but also theoretically has infinite computational accuracy and high replicability. After internalizing past experience in AI models, computational biology can put forward assumptions in an automated, scaled and parallel manner, so that researchers do not need to rely on a few geniuses, while lowering the threshold for downstream development, which is expected to have a significant impact on the industry landscape. The second is to open up a new method of "assumption first - then verification - finally optimization of assumptions", so that R&D efficiency can be improved several times.
As early as 1991, Nature proposed that the starting point of the new biological research method should be that scientists should first start from theoretical inferences, and then return to experiments to track or verify theoretical hypotheses. Computational biology is just able to open up a new way of "hypothesis-verification-optimization hypothesis" based on dry and wet cycle experiments, and improve overall biological research and development efficiency.
Specifically, on the one hand, the laboratory uses high-throughput wet experiments to quickly verify AI predictions while providing AI models with a large amount of available training data to improve the accuracy of the AI prediction model. On the other hand, AI will provide hypotheses (high reference value, even practical) that can be verified in wet experiments based on its own data processing capabilities, and the two will be iteratively accelerated.
3. Precision medicine will become a long-term focus of computing biology
In the field of AI pharmaceuticals, smart laboratories have become an important reflection of the company's long-term competitiveness. The value in application can be divided into three categories according to the process:
1, calculation and deducing biological properties and principles
protein structure prediction, pathogenic mechanism research, protein interaction prediction (PPI), epitope prediction of antibodies and antigen , finding the causes of diseases based on genomics or finding new biomarker , etc.
2, build prediction and judgment model
AI pharmaceutical target-based compound properties prediction (mainly involving small molecule drug development), disease diagnosis/surveillance/treatment modeling, biological simulators covering cells/organs/human bodies, etc.
3, control and transformation of organisms
New therapies/drug development, precision medicine and biomanufacturing (represented by synthetic biology ). Among them, new therapies/drug development is the most mature scenario currently being implemented.Precision medicine will become a long-term focus of computing biology, because the consumption intention in the C-end market is more obvious, and it uses a wide range of human bodies and has relatively direct product forms.
In this direction, there have been many layouts based on multi-omics abroad, while there are relatively few related companies in China, and they are all based on genomics, which has a certain gap.
4. Today's computing biology bottleneck
It can be foreseeable that the future industrial chain of computing biology will be based on data providers as the underlying support + the structural composition of various related practitioners at the upper level (including enterprises that provide computing platforms and software, molecular modeling/machine learning frameworks, computing power and intelligent laboratories).
"In-depth Industry Report on Computational Biology" believes that at present, in order to achieve the above expectations, young computational biology still has the following key bottlenecks to be broken through - some problems are unique to the industry, and some exist in the entire AI science field:
1) Clear the underlying principles of biology
At present, we still have a lot of underlying mechanisms about biology itself to be thoroughly studied. When model construction, biological verification and human body implementation, we need to introduce this knowledge to reduce deviations that do not conform to field cognition and ensure accuracy.
2) Unified computing and data framework
is based on microscopic means, and some specific biological problems can be solved, but to be finally implemented, the required models need to be able to cover multi-omic data, multi-link and function parallelism. In addition, it is necessary to ensure that a variety of heterogeneous data in computational biology, such as images, videos, molecular maps, DNA code, gene expression , electrical signals, etc., have clear standards and general formats to interoperate between different algorithms and platforms.
3) Acquisition of consumer-level data
In the view of analysts, the key industrial development stage of genomics-related computational biology is that data collection has reached the consumer-level level.
4) Project implementation capability
Currently, there are many machine learning algorithms and models in academics that are quite mature. The key is how to add a specific understanding of biology and make detailed adjustments when the underlying data are available. Finally, there is the issue of data privacy and how to make relevant models interpretable and gain the trust of this special industry.
PoS Proof of Stake
cryptocurrencies like Bitcoin require a lot of electricity. In 2021, the Bitcoin network consumed more than 100 terawatt hours, which was more than the annual energy consumption of Finland .
proof of stake provides a way to build a network that does not require much energy consumption. If everything goes as planned, Ethereum, the world's second largest cryptocurrency and running various applications, will transition to this model in the first half of 2022. This shift is expected to reduce energy use by 99.95%.
cryptocurrency runs on the blockchain, and the security of the digital ledger generated through transactions must be guaranteed to prevent cheaters, fraudsters and hackers from invading. Bitcoin and Ethereum currently use proof-of-work algorithms to ensure security: "miners" solve cryptographic problems, thus competing for the right to verify new transaction blocks. Successful “miners” receive cryptocurrency as a reward for their work. Proof of work means finding solutions to mathematical puzzles, which requires a lot of computing power and therefore electricity.
With proof of stake, validators don't have to compete with each other and invest hugely in energy and computing hardware. Instead, their cryptocurrency cache or entitlements allow entry into a sweepstakes. Those selected gain the power to verify a set of transactions (and thus gain more cryptocurrencies). In some networks, validators who exhibit bad behavior are punished and thus lose some of their interests.
Long-term grid energy storage battery
In April 2021, on a sunny afternoon, renewable energy broke the record of California main power grid, providing enough power to meet 94.5% of demand.This moment is hailed as a milestone on the road to low carbonization. But what happens when the sun sets and the breeze stops?
handles fluctuating power production brought by renewable energy to require cheap storage, which lasts for hours or even days, and new iron-based batteries may be able to do this task.
Home-based ESS company in , Oregon, whose batteries can store energy for 4 to 12 hours, has launched its first grid-scale project in 2021. , MA, , MA, raised $240 million in 2021, with its battery capable of storing electricity for up to 100 hours, and its first installation will be at a one-megawatt pilot plant in , Minnesota, which is expected to be completed in 2023.
Both companies choose to use iron-based batteries, and iron is one of the richest materials on the planet. This means that their products may end up being cheaper than other grid storage candidates such as Li-ion battery and Vanadium-based flow batteries.
Form Energy said its battery could end up costing just $20 per kilowatt-hour, even lower than the optimistic forecast for lithium-ion batteries in the coming decades.
However, there are still some challenges to be solved. Iron-based batteries are usually inefficient, which means that a considerable portion of the energy put into it cannot be recovered. In addition, side effects can also degrade the battery over time. But if iron-based batteries can be widely deployed at a low enough cost, they can help more people use renewable energy to power.
AI data generation
In the context of the epidemic becoming the new normal, how to invest in technology has become a concern for corporate managers. At the end of the year, market research firm Gartner will issue a forecast for the "important strategic technology trends" for the coming year, and will inform corporate management, IT practitioners and government personnel to deal with future investment trends and technical risks, while guiding technology and investment directions.
Artificial intelligence (AI) is quite important in the technical trend forecast given by Gartner, involving AI engineering, Hyperautomation, Generative Artificial Intelligence, Autonomic Systems, etc.
Among them, generative artificial intelligence technology ranks first in Gartner technology trend forecast, and is one of the most eye-catching and artificial intelligence technologies. Gartner expects that by 2025, generative AI will account for 10% of all generated data, compared with less than 1% of the current proportion.
so-called generative AI, Gartner explained, learning elements from data through various machine learning (ML) methods to generate brand new, completely original, real artifacts (a product or item or task) that remain similar to the training data rather than copying.
What are the benefits of generative AI? Gao Ting, senior research director of market research firm Gartner, explained that generative AI can not only be judged, but also created. In fact, the biggest use of AI is judgment, which means that the use of AI will have structural changes.
"In the past, we asked the AI to make judgments and categories. For example, AlphaGo, tell me how to move the next step? Ask it to make judgments. Or give a photo to an AI model and say: Help me tell if this is Zhang San, or whether this photo is a cat's photo." Gao Ting told the interface reporter, "We will find that in the future, it often requires AI to no longer make judgments, but say, "Come and help me generate a piece of code. What this code does is add from '1' to '100', and then AI can also automatically generate this code. "
Gao Ting also gave an example that one can use existing data to make a model and then generate more "synthesis data". These synthetic data are like human faces. From the perspective of the naked eye, there is no problem with this face. However, this person is a person who does not exist among the 6 billion people in the world and looks exactly like a real person.
Gartner said that generative AI learns content or objects from data and uses data to generate new and completely original new content, which can next-generation automatic programming, drug development, visual arts, social, commercial services, engineering design and processes. At the same time, it can be used to detect fraud, false information and identity theft. But in addition, although technology companies such as Google , Meta, Microsoft and other must also prevent abuse such as deep fakes. In addition to generative AI, Gartner also pointed out that next year, the trend of AI engineering will also attract industry attention. The engineering of artificial intelligence is to collect data, process data, model, and analyze data, and all reports are processed in SOP ( standard operating program ). The seemingly simple work is of great help to data scientists.
"AI engineering is actually not just a technical problem, it is often a process problem." Gao Ting said that according to statistics, the most time-consuming task of data scientists when processing data engineering is data processing, accounting for 75%, and only 25% of the time left can be defined and solved, greatly reducing the ability of enterprises to solve unfamiliar problems. The latest AI engineering can be integrated into industrial expertise (Domain Know-How). Gartner believes that by 2025, about 10% of companies using AI engineering will be able to achieve more than three times the return rate in their business.
"In 2020 and 2021, the economy has been affected to varying degrees. With the current epidemic becoming a new normal, many CEOs hope that their company's performance will rebound in 2022, or that it can be so-called "wind back" their lost income." Gao Ting quoted a CEO survey report by Gartner, saying that "growth", "digitization" and "efficiency" will be the three keywords for enterprise managers in the coming year. Therefore, the technology trends in the new year are all related to this. In addition to AI technology, new technology trends also include privacy-enhanced computing, cloud-native platforms, etc.
"If the main technical line of last year was 'What changes have happened in the world under the influence of the new crown epidemic', this year's COVID-19 epidemic has almost passed or has become a new normal. How to deal with this new normal, whether it is China or the West, the difference is that everyone deals with it differently." Gao Ting said that under the new normal, working from home has become the mainstream. In this case, companies need to use technical means to recover losses under the epidemic, and how to create a new model under the new normal to ensure the long-term survival of the company has become the main logic of the technology narrative next year.
Malaria Vaccine
RTS, S is the first malaria vaccine approved by the World Health Organization and will be used in children over 5 months of age in medium and high-risk areas where malaria is transmitted in Africa starting from October 2021.
Malaria is one of the three major infectious diseases in the world that seriously endanger human health. With the increasing clinical drug resistance of various antimalarial drugs such as artemisinin, nearly half of the world's population is still at risk of malaria infection, and the most fatal malaria protozoa falciparum causes 200 million to 300 million infections every year. Since the 21st century, about 10 malaria vaccine projects around the world have been approved for clinical trials every year, and about 150 clinical trials have been completed or terminated early.
HTM0So far, the malaria vaccine RTS, S is the only vaccine that has been proven to reduce the clinical morbidity and mortality of children with malaria. The malaria vaccine RTS, S only has a high protection rate for children aged 5-17 months within one year after 4 doses. The immune protection rate then dropped rapidly, and the average protection rate after one and a half years of vaccination has been lower than 30%. As a zero breakthrough in the field of malaria vaccine research, the malaria vaccine RTS, S has great practical significance. The World Health Organization expects to save the lives of tens of thousands of African children under 5 years of age each year in the future.
malaria vaccine RTS,S has not met the official standards of the World Health Organization, that is, the protection rate is greater than 50%, and the protection time is greater than one year. Therefore, how to effectively curb the epidemic and spread of malaria in countries and regions such as tropical and subtropical countries is still a scientific problem that global malaria researchers urgently need to solve.
Because the life history of Plasmodium includes complex growth periods such as the liver (cell) internal phase, red (cell) internal phase and mosquito phase, Plasmodium falciparum has highly mutated antigen proteins and variable immune escape strategies, which not only limits the research and development of malaria vaccines at home and abroad, but also leads to the main reason for the imperfection of malaria vaccines RTS and S.
In recent years, with the widespread application of a variety of new gene editing technologies in the identification of key biomarkers of Plasmodium falciparum, it has become possible for researchers to design multivalent vaccines for different growth periods of Plasmodium falciparum. At the same time, compared with traditional vaccines, the technological innovations of emerging messenger ribonucleic acid vaccine technology, vaccine adjuvants and antigen delivery systems will also provide more potential solutions for malaria vaccine research, making the research and development of a new generation of highly efficient malaria vaccines expected to achieve key breakthroughs in the next 5-10 years.
Carbon removal factory
Since the industrial revolution, human activities have emitted a large amount of greenhouse gases such as carbon dioxide (CO₂), and the greenhouse effect has continued to strengthen, resulting in a continuous increase in the global average temperature.
In fact, even if the whole world has reached carbon neutrality, since humans have emitted more than one trillion tons of CO₂ since the Industrial Revolution, it would be a very slow process to reduce the CO₂ concentration in the atmosphere to the level before the Industrial Revolution.
As a technology that uses engineering systems to remove CO₂ from the atmosphere, the large-scale application of Direct Air Capture (DAC) technology is of great significance to effectively reduce CO₂ concentration in the atmosphere and curb climate change. This technology mainly uses a induced fan to pump air in, capture CO₂ through adsorption, absorption or membrane separation device, and discharge CO₂ back to the atmosphere. The captured CO₂ can be sealed or utilized. The entire process can be understood as an industrial "photosynthesis".
Unlike CO₂ capture technology for industrial fixed sources, DAC technology can be deployed anywhere in the world with electricity supply, with more flexible location selection and modular construction.
DAC technology has obvious technical advantages in carbon removal, but the current high operating costs are still a key factor limiting its large-scale application. Recently, researchers at the University of California, Berkeley have made a prospect of its development and proposed a policy roadmap suitable for the development of this technology. They believe that the global promotion of DAC technology cannot rely on market leverage effect, but should promote its large-scale deployment through the continuous "fiscal incentive + mandatory deployment" policy. From a technical perspective, the key to the development of DAC technology lies in the research and development of efficient and low-cost carbon capture materials and process systems, and its commercial application still needs to rely on technological progress to significantly reduce operating costs.
In recent years, developed countries in Europe and the United States have successively carried out the research and development and application of DAC technology, and continuously reduced operating costs through the advancement of materials and technology. In August 2021, the U.S. Department of Energy announced a $24 million allocation to support DAC technology, and some "carbon removal plants" larger than the carbon dioxide capture plant Orca are also under construction. These advanced work may enable developed countries to master cutting-edge technologies and core intellectual property rights earlier and seize the initiative for future economic benefits.
COVID-19 mutation tracking
COVID-19 virus is still spreading around the world. Among all nasal swabs that have positive reactions to COVID-19, about 200 have been sent to gene sequencing machines for additional analysis. The purpose of this is to create a new map for the genome of the SARS-CoV-2 virus and see what changes are happening. This map consists of about 30,000 letters. Gene monitoring such as
allows scientists to quickly discover and warn of new variants such as alpha (a), delta (δ)
and most recently Omicron. This is an unprecedented work that makes SARS-CoV-2 the most sequenced organism in history, surpassing influenza, HIV, and even our own human genome. Open databases like GISAID and Nextstrain have shown gene maps of over 7 million bacteria.
Omicron is the most varied variant so far.In November 2021, a South African laboratory discovered a viral genome with more than 50 variants in its sequencer and issued a warning signal for the first time. Almost instantly, computers in Seattle, Boston and London were using this data to predict: Omicron is a trouble, a variant that could evade antibodies. One thing the
sequencer can't tell us is how exactly will SARS-CoV-2 evolve next. That's why some say we should track this virus more closely. Most of the sequences are generated in the UK, the United States, and Denmark, but in areas without sequencing, the virus can still evolve without realizing it. Fortunately, South Africa’s rapid work on discovering Omicron and tracking its spread provides early warnings for the world.
Blue Ocean Brain Life Science Solutions
1. Case Overview
A certain pharmacy school was initially founded in 2021. Drug molecular research and development is a very complex and time-consuming process. Drug molecular screening is just a link in the early process. For example, if you look for small molecules that bind to protein virus enzymes, because there are ligand (small molecule) libraries of different types or research institutions, the number of ligands (small molecule) libraries is huge, and the number of ligands in each ligand library is tens of thousands (or even larger), so it is unrealistic to test and verify through experimental methods. Through computer numerical simulation, the binding effect of different ligands is scored, and some ligands with high scores and reasonable binding mode are screened for experimental verification as candidate drugs, which can effectively accelerate the drug development process.
Due to the huge number of ligand libraries, the screening is also huge. For example, the ligand library has 10,000 candidate ligands, with an average treatment time of 1.5 hours per ligand, which takes a total of 15,000 hours (625 days). Therefore, in order to complete the calculation within the specified time, the following conditions are required:
1) a computing platform with strong computing power,
2) a large capacity storage, used to store processing data and calculation results
. In addition, in order to ensure that the screening calculation can be completed efficiently and smoothly, computing services are also required, including:
1) a cluster software operation environment, ensuring that it runs under multi-machine environment software, and data access
2) a parallel solution that can support multi-task concurrent processing in multi-machine environment
In addition to the computing platform, drug screening also requires high-performance application software. Drug screening simulation calculations include Docking and molecular dynamics calculations: Docking takes relatively little time and is often used for preliminary screening of a large number of ligands. The main software includes dock6, Autodock Vina, Glide, etc. The molecular dynamics simulation calculation is time-consuming and the time changes in the test effect are used to further analyze the results of Docking primary selection. The main software includes Gromacs, Namd, Amber, etc. The acceleration effect using GPU is generally obvious.
2. Solutions and Values
Drug Small Molecules Research and Development requires a high-performance cluster with powerful computing power. Obtaining these computing resources and services has become the top priority at the moment. The School of Pharmacy of Tsinghua University has built four A100 liquid-cooled servers, 9 CPU servers, and two high-throughput liquid-cooled servers to provide a basic computing platform for a high-performance computing environment.
When using DOCK6 to process docking cases of ligand (small molecule) libraries, in a folder, such as mol2, stores a large number of small molecule files. Each small molecule processing process is the same and needs to be calculated with the same receptor (such as viral protease). At this time, a large number of GPU and supercomputer products are required, as well as all-weather technical support in a short period of time. Build an open sharing platform and use high-performance computing clusters to use molecular docking, molecular dynamics simulation, and deep learning model training for drug research and development, shorten the calculation work that takes several days to several hours, and increase the speed by 8 to 20 times. At the same time, different sub-accounts are created for each R&D teacher to realize computing resource sharing and data sharing. A basic computing platform for high-performance computing environments is also needed to achieve efficient drug screening.
3. Summary
This school of medicine research and development requires a high-performance computing cluster with strong computing power, such as drug screening, which requires docking of a large number of small molecules. Teachers from the School of Pharmacy can use Blue Ocean Brain’s high-performance liquid-cooled server to quickly build high-performance clusters, obtain high-performance computing instances, and meet the needs of computing power. At the same time, it provides a solution for high-throughput task processing, so that drug screening is processed concurrently on multiple computing nodes and multiple cores, reducing the overall task execution time.
