Recently, Ali cloud announced its container service scheduling GPU accelerated computing cloud server starts, the fastest contrast only 60 seconds to complete the work of the new crown virus nucleic acid; at the same time will be medical research institutions, research institutions, CDC and other front-line virus free Open gene computing services, technology can significantly improve metagenomic sequencing, vaccine research and development related to processing efficiency.
Based on this, the grain less special trip interview with Gene computing Ali cloud services AGS person in charge, senior technical expert Li Peng, centralized presentation for the GPU and container technology dramatically nucleic acid than about the details of speed and a lot about Ali cloud gene Computing Services (AGS) of information.
It is crystal little is understood that this gene Ali cloud computing services AGS virus than the ability -mNGS metagenome comparison from design, development, verification to the final on the line took only 20 days, but Li where the team was entrusted with the arduous day work. Critical starting point is that, in addition to providing faster than the ability of the data for the new crown pneumonia virus mNGS metagenomic sequencing, it is more important to give research institutions more focused on business innovation, rather than the data itself repetitive processing work, which it is related to the type of biological computing solutions has the distinct particularity.
Crystal less: Biological computing solutions have a certain degree of particularity, compared to other types of technical solutions What's the difference?
Gene Ali cloud computing services AGS person in charge, senior technical expert Li Peng: Unlike conventional computing, genetic computing there are high demands on computing and massive data storage capacity, mainly through the container calculated auto-scaling features and Ali cloud ECS automatic scalability the open, this allows AGS have the ability to scale computing resources on schedule cloud. Specifically, after a reasonable segmentation data gene specific algorithm, can achieve large-scale parallel sample data processing and computing TB level.
Gene calculate past service to clustering, hybrid cloud, elastic AGS-based workflow service. But in the two years through the years, Ali cloud computing solutions provide a genetic hybrid cloud-based container, the overall technology is cloud native container under the hybrid cloud infrastructure, cloud reached under the cloud resource integration, unified management of cross-regional clusters. In 2020, in the conventional technique, based on precipitation, Ali cloud upgrade the capacity of the AGS, the use of public cloud services API to do that is ready to use, run out and go to this virus detection, for example, in providing custom virus database on the basis of flexibility and make calls without the service of, the first call to final data that is able to output ratio and return of results to the user.
Real-time quantitative PCR (RT-PCR) and sequencing of pathogenic macro (mNGS) is a method for early diagnosis of new viral infections crown. As the two technologies have advantages and disadvantages, in combination in order to achieve complementarity.
RT-PCR is the new crown main viral nucleic acid detection methods, relatively low cost, simple operation. There are now many new crown virus RT-PCR kit is optional, but due to virus concentration and kits quality, high false negative result occurs, leading doctors and patients often need to be repeated several times to detect and long waits for test results.
mNGS technical advantages will be able to be detected by the primary investigation of all known pathogens, effectively avoid the trouble of repeated sampling and detection times required for screening large numbers of samples difficult to achieve the problem. When the virus spread through several generations, will mutate, mNGS is more effective tracking solution, with the virus can accurately determine the virus intergenerational relations, and even hair to find the source of the virus. Based on the analysis mNGS sequence alignment manner, once the genome of a pathogen known, by updating the database, it is possible to achieve an efficient and accurate detection of virus similar functions, and other pathogens while mixing other than the new integrated viral infection identified. But the disadvantage is the higher cost of mNGS need sequencer and professional Seng Credit analysis, and complex operation, the detection time is relatively long, and therefore can not achieve large-scale promotion and rapid diagnosis.
Today, through the use of container Ali cloud services, researchers simply on Ali cloud line AGS rna-mapping API command to start the alignment process, then Ali cloud container service scheduling GPU accelerated computing cloud server starts, you can get high quality in 60 seconds and preliminary data matching quality report reads, providing fast and accurate data support for multiple pathogen detection, and further research and study new variation crown viral proteins.
Based on this, we found that this cloud Ali cloud service providing significantly different from the past, "full set" thought running through them, that is used by organizations no longer need to purchase additional computing resources, not to care about how to build a cluster algorithm and consider you can call the network and many other issues, simply store the data into the object and upload, wait for the authorization to complete the cloud service to process the data, and output the results to generate relevant reports, "not only that the use, throw-away" outstanding qualities.
In addition AGS has the fastest acceleration than the cloud currently in process under machine sequencer data comparison, sorting, mutation testing and other areas, you can reach a hundred times ordinary CPU processing speed, while maintaining BWA / GATK other mainstream gold standard software the precision and recall, a significant reduction in the cost calculations, cost is very impressive, also shows the deep exploration of its technical skills, but what exactly is the opportunity to contribute to a spark between the container computing technology and gene burst?
Crystal little: when and what an opportunity to begin to accelerate gene discovery related to the calculation made for container technology? This API call more what means?
Gene Ali cloud computing services AGS person in charge, senior technical expert Li Peng: We observed that the container in one hand, data striping technology, quality control, Call variation provides standardized processes, acceleration, elasticity, authentication, observation, measurement and other capabilities ; on the other hand, need the help of high-value container mining machine-learning platforms and parallel computing framework for the completion of large-scale linear algebraic model for precise medical care. Over the past two years, AGS has been in the industry through clustering, hybrid cloud, flexible workflow-based AGS way to help service companies to complete large-scale gene sequencing day handle more than 100TB of data, 1-year deal with dozens of PB data tasks, with remarkable results.
And this API of trying to do can effectively help users get rid of the complexity of the cluster resource creation and maintenance costs, storage costs with very little can be done any number of on-demand computing needs, costs only in accordance with the data processing and the computing capacity to pay, a real no service and pay as you go.
There are news that Ali cloud as early stages of the epidemic severe announced to provide all the AI crown to the new operator force free vaccine development research institutions, has received the Zhejiang University School of Medicine, GHDDI global health drug development center, including Tsinghua University School of Medicine, Beijing dozens of authoritative scientific research institutions at home and abroad such as applying University School of Medicine, the first operator to apply force resources have been deployed on the line, continue to overweight to speed vaccine development, but in one case specifically, how do?
Crystal few: the current count of how to apply the external force to deploy the whole situation? What effect? What specific aspects of the follow-up support services there? What are the future plans in booster vaccine in development of drugs?
Gene Ali cloud computing services AGS person in charge, senior technical expert Li: current AGS has launched domestic services in various districts, especially in Beijing, Shenzhen, Hangzhou and other companies have started production using large-scale gene; additional foreign multiple regions also opened a service, such as Frankfurt, the hardest hit by the epidemic may also have access to the AGS service.
Crystal less: the new coronavirus outbreak, in order to help anti-epidemic attack, Ali cloud provides free high-performance computing to the global public research institutions, SCC supercomputing clusters and CPU / GPU machine, cloud supercomputing and AI techniques, including help in E-HPC cloud ultra-count drug R & D, how this relationship with AGS? AGS occupies what position in the external output of Ali cloud technology or how to play the role?
Gene Ali cloud computing services AGS person in charge, senior technical expert Li Peng: AGS mainly located in the Serverless public cloud services out of the box, provided to each partner ecology, such as genomics companies, research institutions, schools and other PaaS services, it can be integrated in a variety of cloud SaaS applications to achieve efficiency hundredfold improvement and significantly reduced costs. Of course, E-HPC is also a great advantage Ali cloud product lines, mainly for high-performance computing research institutions to make polymer analysis, of course, scientific research institutions, the CDC can use both AGS and E-HPC to form a complementary research projects, and Do not conflict.
The interview drew to a close, one can imagine that the future of computing will continue to move AGS gene Serverless and pay by volume, elasticity of demand continues to be the development of an integrated course, have the opportunity to provide accelerated processing of various genes of the human genome, as well as animal and plant viruses cloud service. Overall, with the cost of sequencing, calculate costs and declining cost analysis, there is reason to believe that gene sequencing and accurate medical in the near future, ordinary people will go to the list of routine health services, human perception of their own genes deeply bound to promote and enhance life might overcome various diseases, cloud computing services AGS will promote the country and the world in the Pratt & Whitney accurate medical early arrival.
Under Contagion, exactly what are Black & urgently to be discovered? We just wait and see more!
