Demand for life science research drives cloud computing

The rapid growth of life science research data has led the industry to turn its attention to cloud computing applications, especially in the field of genetic research. Chris Dagdigian, Principal Consultant at BioTeam, said that the use of nano- and microfluidic chemistry has enabled the use of camera sensors to monitor the DNA on tiny particles. Analyze TIFF image data with a data volume of 800 gigabytes. “This raises a lot of data capture and control issues. We are currently in an era where the device can be generated in a small wet lab, but it can generate 1 terabyte or more of data per day.”

Eric Schadt, executive director of genetics at Rosetta Inpharmatics, agrees. "The next generation of gene expression and DNA sequence technology is generating large amounts of data faster. We believe that microarrays and high-density SNP arrays are generating high-dimensional and difficult-to-master data. The next generation of technology will be of the current one to two. Times.

Schadt said the data should answer how the complex disease system manifests itself in the human body. Not just DNA mutations will tell us how genes are linked to a disease. More, we will see people integrating DNA mutation information with gene expression information or metabolic information or protein information. Researchers will configure the cloud computing system to get all relevant data information.

Although Merck has not yet configured cloud computing. The company built a computer center in Rosetta, Seattle, which was built in 2001 with approximately 10,000 processors and a network-based fine structure that allows Merck's researchers in any area and thousands of projects to access storage. system data. But this situation is about to change.

In the coming week, Merck will install the Rosetta computer group and store most of the data there in the non-profit life science information analysis database Sage Bionetwork, which is led by Schadt and Stephen Friend (Merck Research Laboratories Senior Vice President and Oncology) The person in charge) built it. The pharmaceutical company will open its doors into Sage and will integrate its research computing system for the New Center (Excelence for Molecular Profiling & Research Informatics). At the same time, Sage will work with other public and private research centers to expand the database. According to Schadt, with the development of Sage, the cloud system will be further integrated.

As pharmaceutical companies began to adopt cloud computing systems, service providers built distributed computer equipment structures to match demand growth. Amazon saw a rapid increase in demand for its cloud storage service (Amazon EC2, which went on sale in 2006). According to information provided by Amazon Web Services Vice President Adam Selipsky, the number of server files has increased from 18 billion to 52 billion in the past year.

Selipsky said that Amazon's cloud services can be viewed as virtual data services with flexible and near-infinite capacity. “This service is much like the physical equipment of a company's data center, just in our plant and not in your data center.” And it is because any company can add software and personal applications to the underlying computing infrastructure, cloud systems. Users run their own applications in the distributed structure of Amazon over the network. According to Selipsky, Amazon offers pay-per-use storage, information alignment, and high-performance computing services.

Selipsky notes that this service has generated great interest from life science companies, and the data processing pressure of these budget-bound companies is growing. The speed of data processing is also the reason why researchers in pharmaceutical companies are turning to cloud computing. He added that Amazon will introduce Amazon Elastic MapReduce, a Hadoop-based cloud computing device, from the Apache Software Foundation's web application that speeds up processing by allowing computers to process thousands of files and petabytes simultaneously. 1 petabyte is equal to 1,024 tetabyte or about 1 million gigabyte.

Supplier perspective

Mike Naimoli, director of life sciences at Microsoft, said that in addition to time and cost savings, cloud computing can also share data and connect with research organizations and other partners. The company's Azure Services Platform service platform for cloud computing was launched last year.

"This platform is used to collect data, connect and process data. It implements its data collection functions through applications running on Azure. Microsoft does not develop such applications. We provide user applications that can run user applications," he said. Basic framework and structure. Any user can run the cloud system locally or over the network."

Rishi Chandra, senior product manager for Google's cloud computing service Google App Engine, said the urgent need for data storage and computing will prompt pharmaceutical researchers to adopt cloud services. "Cloud services in a distributed network run more meaningful because you can control more of the resources you need."

Chandra said that Google allows users to place data from their cloud system structure behind a secure firewall. The management of data access requires some customized integrated services. Although Google and Amazon and Microsoft are similar to each other, managing the security of their physical computer structure, users will be responsible for data encryption, data access management and other cloud system security measures.

Part of this work was done by third-party software vendors such as Cycle Computing, which began developing open source software for high-performance computing systems four years ago. Jason Stowe, CEO of Cycle Computing, said the company has now brought commercial software and security management services for cloud computing systems to market.

The company has a partnership with Schrödinger, a computational chemistry software company. Stowe said Cycle is developing applications for the next generation of genome sequences that enable researchers to use cloud computing to process and streamline lab raw data.

Cloud computing is a brand new field

While users and vendors agree that these are just the early days of cloud computing products and services, they see this IT service as a new choice for data storage and processing, and the service offers a host of other services. Some users see it as a secure environment for collaborative environments and clinical trial data.

Karen Riley, a spokesperson for the Food and Drug Association, said cloud computing is a whole new field. "If cloud services become the storage center for clinical trial data, our focus will be on system write security to avoid data tampering," she said. "The audit firm (such as Google) has no experience, and the trial investors will still Take responsibility for data security maintenance. Riley noted that the FDA trusts secure external servers for clinical trial data email transmission.

Powers said that we are now in the mouth, which is exciting. Current Lilly cloud computing applications have led to a lack of state change in writing research and clinical trial management. He said that the cost savings and speed of research brought by cloud computing make this service more attractive to pharmaceutical companies, so the potential market for cloud computing services is huge.

“We changed the way we do business. More work is done collaboratively. As we move forward, infrastructure pressures increase and we need to increase our unfamiliar IT infrastructure. The current problem is that we are Is building infrastructure still for cloud computing services?"