To Improve Cancer Treatment, Look to the Sky

Completed in 2003, The Human Genome Project was tasked with sequencing the chemical base pairs which make up DNA, and with the identification and mapping of 20,000 to 25,000 genes of the human genome. The objective of the project was for researchers to find disease-causing genes and locate patterns in gene expression which could help physicians develop more specific treatments in fighting disease. One area of particular research interest is sequencing the genetic variations in the BRCA2 gene. BRCA2 is a gene on chromosome 13 that normally helps to suppress cell growth. Researchers are studying mutations in the BRCA2 gene which have been linked to a higher risk of breast, ovarian and prostate cancers. While genetic mapping of the human genome has far-reaching treatment implications, it has been met with limitations in data processing.

Genomic studies and DNA sequencing are massive computational challenges. These clinical analysis techniques require extremely high levels of data throughput in the form of powerful storage and computing platforms.  For instance, the first genome sequence took up about 750 gigabytes of data that had to be analyzed, backed up, and archived for long term storage. Today, scientists and researchers engaged in genome studies often work with calculations so complex that it could take years for individual computers to complete them. However, these types of studies, in which hundreds of human genomes are sequenced, are now becoming feasible.

So, how can technology meet these throughput challenges? The answer is cloud-computing. GlobalData believes by leveraging the power of cloud computing, researchers will have the processing capabilities needed to make sense of the incredible volumes of data produced by genomic studies, and to deliver actionable insights to improve cancer treatment. While cloud, or grid computing, is an emerging field of computer science, consumers and multinational corporations have been benefiting from cloud usage for a number of years. For example, Apple’s iCloud stores consumers’ music, photos, and documents wirelessly and streams this information to a user’s PC or mobile device. The benefit is that the software and storage do not exist on the user’s computer or hand-held device - Apple’s back-end grid of networked computers that make up its iCloud handles the workload. Similarly, through grid-computing, geneticists and oncology researchers can now send the data and calculations from genome sequencing to the cloud for processing. The cloud system taps into the processing power of all available computers on a vendor’s back-end, significantly speeding up the calculations.

Beyond the research implications, the use of cloud-computing could also benefit patient care. In terms of cancer treatment, specific gene-level changes in cancer cells drive critical treatment decisions. In the past, doctors’ offices had to send patients’ DNA samples to a contract lab for sequencing and analysis, which could take weeks, wasting valuable time which could otherwise be spent on treatment. However, with disease-specific and patient data in the cloud, a hospital’s onsite medical staff would be able to analyze the data in a matter of minutes, allowing the staff to tailor a patient’s treatment to his or her specific tumor. This tool would also provide for detection of the genetic conditions that make some people susceptible to certain types of cancers.

One vendor illustrating the convergence of cloud-computing and genomics research is Knome, a life sciences company specializing in developing software tools for genomic sequencing. In June, Knome released kGAP 2.0, a cloud-based genome informatics software engine built to automate the process of finding genetic variants that influence the risk profile of developing cancer, its progression and its treatment response. By leveraging Knome’s data cloud, kGAP 2.0 can simultaneously annotate and interpret hundreds of genomes from multiple sequencing platforms from diverse data sources into a single genomic study, completing in a day what would otherwise require months to compile.

In addition to Knome, Amazon Web Services (AWS) and the United States National Institutes of Health (NIH) announced that the complete 1000 Genomes Project data is now accessible on AWS as a publicly available data set. The data set is stored and accessible on the company’s Simple Storage Service (S3) and Elastic Block Store (EBS) public cloud services. The 1000 Genomes Project is an international research effort comprised of 75 companies and research institutions coming together to establish the most detailed catalogue of human genetic material. The project, which started in 2008, has grown to 200 terabytes of anonymous genomic data including DNA sequenced from more than 1,700 individuals. This makes one of the largest collections of human genetics data publicly available to researchers to collaborate worldwide to study diseases, such as cancer.

This is a rich opportunity for geneticists and cancer researchers who want to design large-scale genomic studies, but previously lacked the computational and storage capacity. Prior to the data being put in the AWS cloud, researchers who wanted access to these public data sets had to download them from government data centers to their own systems, or have the data physically shipped to them on discs. This was impractical - the 200 terabytes of genetic data is comparable to 16 million file cabinets of text or more than 30,000 standard DVDs – most research labs do not have this computing and storage bandwidth internally to run sufficient analyses. Now, with the public data sets placed in the AWS cloud, researchers and labs of all sizes and budgets have a simple way to access the 1000 Genomes Project data and can start analyzing subsets of the data without the added financial investment that would normally be required in provisioning the hardware necessary for such cancer studies. We believe that cloud-computing will accelerate the pace of new genomic discoveries by creating an open ecosystem allowing more investigators access to these important data sets.

Beyond companies such as Amazon and Knome, large hardware vendors are partnering with research institutions to improve cancer treatment. Last November, Dell announced it donated its high-performance cloud computing resources to research treatments for pediatric cancer. The company has gifted its computing muscle and data exchange capabilities to the Neuroblastoma and Medulloblastoma Translational Research Consortium (NMTRC) and the Translational Genomics Research Institute (TGen). The gifted cloud is powered by Dell PowerEdge Blade Servers, PowerVault Storage Arrays, and Dell Force10 Network infrastructure. GlobalData believes that the goal of the partnership is to provide cancer researchers with the computational resources to perform complex analysis on cancer patient genomics to develop treatments faster, so that personalized therapies can be better targeted to individuals in hopes of improving clinical outcomes. These institutions will use the donated assets to conduct the world’s first personalized medicine clinical trial for neuroblastoma.