|Focus on: Microsoft Genomics|
Genomics research is playing a foundational role in the effort to revolutionize how health is improved and diseases are treated. Until the advent of the cloud, however, sharing genomic data was cumbersome and costly, which stood in the way of its widespread application.
With the help of our partner, DNAnexus, St. Jude Children’s Research Hospital recently leveraged the Microsoft Genomics service to develop St. Jude Cloud. We sat down with Jinghui Zhang, PhD, St. Jude Chair of Computational Biology, and St. Jude CIO Keith Perry to discuss how St. Jude Cloud is advancing the global fight against childhood cancer, supporting data sharing, and encouraging greater collaboration among scientists worldwide.
What does pediatric cancer research look like today?
In pediatric cancer research, there are three major sub-categories: leukemia, solid tumors, and brain tumors. Leukemia was the first type St. Jude studied, and we’ve made dramatic improvements in terms of treating it. In 1962, when St. Jude opened, the cure rate was only 4%, and now St. Jude is achieving a 94% cure rate. However, cure rates are still below 80% in other areas, and optimizing treatment protocols in those areas has reached a plateau. We need to develop new approaches to improve these cure rates and reduce the risk of mortality and relapse.
The reason we do genomic research is two-fold. First, we can improve patient diagnoses and tailor treatments accordingly. And second, we can continue to optimize therapies in a way that gives patients high quality of life as survivors. We are trying to discover how genomics can be used as an analytical solution to these problems.
How important is collaboration to this research, and how do researchers collaborate beyond organizational walls?
Collaboration is extremely important. Pediatric cancer is less common than adult cancer, and genomic findings overlap by only 50% between the two. This means there are fewer subsets of data to work with. For example, if you want to accumulate 500 breast cancer specimens, it’s easier to do because the incidence of adult cancer is much higher. In the pediatric cancer community, some cancer types are so rare, there is no way to accumulate sufficient samples to identify patterns without collaborating – you might even have to reach out at an international level.
What are some challenges cancer researchers face in using genomic data?
There are general challenges for any organization that wants to use genomic data, such as handling and accessing large datasets. Often, these datasets are accessible only to large institutions that have the supporting computation infrastructure, which is something smaller, individual laboratories don’t usually have. In our experience, troubleshooting issues with individual deployments of genome sequencing tools or systems has also proven problematic. Overcoming any of these challenges is extremely time-consuming and costly, and that stands in the way of discovery and advancement.
You’ve recently collaborated with DNAnexus to build St. Jude Cloud using the Microsoft Genomic service – can you tell us a little about St. Jude Cloud?
St. Jude Cloud aims to break down barriers to entry and create a collaborative research space. Scientific discovery, especially in the field of cancer genomics, requires sophisticated technology. By building our system on cloud computing infrastructure, we can install once and everyone can access it immediately. You don’t need to have a high-performance computer at your academic center to research and leverage St. Jude Cloud, and you don’t need to download tools. You can use the tools we’ve already validated and developed to run on data you’re producing or on other data you want to analyze.
What motivated St. Jude to develop St. Jude Cloud?
Sharing research and scientific discoveries is vital to advancing cures and saving more lives, especially when it comes to rare diseases like pediatric cancer. We have first-hand experience suffering through the laboriousness of downloading public data and seeing the limitations of genomic clouds in supporting end-to-end workflows and visualizations. We’re one of the largest generators of genomic data, so at some point the question became, if we don’t improve this process, who will? That said, we are not just the developer of this process – we’re also users. We wanted to make it easy to use and consume data and enable users to avoid the frustrations we experienced.
How does St. Jude Cloud help researchers overcome challenges in using genomic data?
St. Jude Cloud enables researchers to overcome the bottleneck of data accessibility – even individual laboratories of computer scientists or bench scientists can access and use our genomic data and tools with the opportunity to develop new ones. This type of contribution is mutually beneficial to the individual researcher, whose data and tools gain more visibility, and to the community, whose pool of resources grows. It truly takes a community to make full use of the genomic data and advance pediatric cancer research, so we hope our investment will benefit everyone involved.
What were some of the characteristics you looked for when you started exploring cloud platforms and partners?
The ability to protect patient confidentiality was top of mind. On top of using this infrastructure for the research purposes we developed it for, we envisioned that there could one day be a clinical application built on top of it, so data security was of utmost importance. We needed an infrastructure with a good track record in genomic data analysis, workflow development, and data protection, which made DNAnexus stand out when we were searching for a partner.
What unique attributes did Microsoft bring to this partnership?
We partnered with Microsoft because they took time to understand us and our mission. Our goal is different from most organizations’ – we want to get our data onto a platform where people can collaborate effectively and accelerate discovery.
Who will benefit from St. Jude Cloud?
St. Jude Cloud has a very strong emphasis on supporting regular scientists – you don’t need a PhD in computer science to use it. It is not only helpful for cancer geneticists and computer scientists, but for clinicians, post-docs, and oncologists who may not have a genomics background. We always say our end goal is to make genomic analysis as easy as filling an online shopping cart.
Now that the foundation has been laid for this type of collaborative research, what’s next?
Once you put data in the cloud, you can start having meaningful conversations. Now, it’s not a question about moving data – it’s about what you want to do with it. We believe the applications of this research go beyond cancer cells – even research on non-cancer-based genetic diseases can benefit.
Discover more about St. Jude Cloud and the Microsoft Genomics service
You can learn more about how St. Jude is enabling collaboration and accelerating scientific discovery here. Or, explore how Microsoft Genomics can power genome sequencing and research insights for your organization.