Patients don’t want to wait months for potentially life-changing genetic testing results. How could Blueprint Genetics deliver accurate results, fast?
Cognitive analytics lets Blueprint mine vast bodies of medical literature and present results in a structured workflow, helping to identify potentially deleterious variants and build reports quickly.
14days average turnaround time for test results, when competitors can take over a month
20xexpansion of product portfolio, from 20 tests to 400, in just seven months
Up to 50%rise in productivity with automatic cognitive analysis of medical literature
Business challenge story
Breaking new ground
Blueprint Genetics was founded in 2011 with a vision of disrupting the genetic testing industry by harnessing leading-edge information technology and robotics to automate the sequencing process and handle unprecedented volumes of tests and samples. Its founders’ ultimate goal was to provide a one-stop shop for screening for the whole field of inherited disorders—over 3,000 types of disease in more than a dozen medical categories—and deliver results within weeks rather than months.
Samuel Myllykangas, CTO at Blueprint Genetics, explains: “We decided to start with a single category: a set of 20 tests for inherited cardiovascular diseases. After four years, we had proved that our business model was successful, and we wanted to broaden our operations to support a much more comprehensive range of screening services. This meant we needed to scale our product offering by a factor of 20, expanding from 20 tests to an offering comprising more than 400.”
The company reviewed its core business processes to identify the best way to scale up its operations.
“Our work involves receiving blood samples, performing the genetic sequencing in our labs, processing the output to turn it into human-readable data, and then interpreting that data and reporting our findings,” says Samuel Myllykangas.
“We realized that we could easily ramp up the sequencing stage by using the robotics in our lab, and accelerate the data processing stage by investing in more computing power. However, the final interpretation and reporting stage was much more difficult to scale, because it relies so heavily on the expertise of human analysts.”
To interpret the data, these analysts had to look at each unusual gene or mutation, and then look up case histories and medical journal articles in a range of different databases to identify genotype/phenotype combinations acting as markers of particular diseases or disorders. Recognizing these combinations required a lot of domain expertise and experience, and because of the rapid rate at which new medical research is published, there was always a risk that a relevant article or database entry might get overlooked.
“Essentially, the challenge was to find a way to simplify, structure and automate as much of the analysts’ work as possible, to allow them to interpret the data and create the reports more quickly and easily,” remarks Samuel Myllykangas. “We needed something like an ERP system for genetic sequencing—but as far as we were aware, nobody had ever tried to build such a system before. We realized that we needed to break new ground and design a solution for ourselves.”
Finding the right way forward
To help develop its new system, Blueprint Genetics consulted IBM.
“We realized that the volume and variety of information involved in interpreting genetic sequencing results meant that this was fundamentally a big data problem,” explains Samuel Myllykangas. “We didn’t have any expertise in that area, so we looked for a partner to help us design the architecture for a solution. IBM had the most complete proposal, encompassing both the data exploration and process management capabilities that we needed.
“The IBM team also helped us to structure the project and set sensible goals, as well as introducing us to two excellent partners here in Finland—Elinar and Enfo Rongo—who helped us with the development and implementation of the solution itself.”
The solution that Blueprint Genetics has developed consists of two equally important parts. The first is a big data analytics solution that harnesses IBM® BigInsights® for Apache Hadoop to ingest, enrich and store the data generated by Blueprint’s genetic sequencing engine in a Hadoop cluster. Each sample generates 20 or more separate files of different data types, including text, images and structured data, which are combined together into a single record.
IBM BigInsights Text Analytics is then used to analyze the abstracts of articles from PubMed and other external medical databases and identify any articles that mention the specific combination of genes or mutations in the sample. This cuts the number of articles that Blueprint’s analysts need to consider from thousands to just the top ten or so, ranked by relevance.
IBM Watson™ Explorer acts as a flexible front end for the solution, making it easy for the analysts to find and explore these articles and compare them with the sequencing data via a single, intuitive user interface. This eliminates the need to log into multiple different databases to gather information, hence saving time and increasing productivity.
To deliver this part of the solution, Blueprint worked with Enfo Rongo, an IBM Premier Business Partner with deep expertise in big data technologies.
Samuel Myllykangas comments: “Enfo Rongo did an excellent job of taking the concept we had developed and implementing it successfully in a real-world environment. We understand that there are not many partners in the region who have the competence in big data technologies to deliver such a solution.”
Tommi Salmi, Director of Enfo Rongo, adds: “The project was a learning experience for both parties, and we’re proud to have proven our competence and delivered a solution that gives Blueprint’s analysts the insight they need.”
The second part of the solution is a comprehensive workflow management system that structures the interpretation process, guides the analysts to work in a standardized and efficient manner, and pre-populates the final reports with as much data as possible to reduce repetitive manual work.
To develop this workflow management component, Blueprint worked with Elinar, an IBM Premier Business Partner that specializes in enterprise content management and knowledge management solutions. The case management capabilities in IBM Business Automation Workflow are used as the engine to drive the workflows, so that when a new sample has been analyzed, it is automatically assigned to the appropriate analyst. The sequencing data is then filtered twice: first to focus on the genes that need to be tested, and then to classify the analyst’s findings. Clinically relevant findings are automatically added to a reporting template, and the analyst can edit it and add commentary. Finally, the report passes through an approval workflow to check that everything has been completed correctly before it is sent to the customer.
Ari Juntunen, CTO at Elinar, comments: “The Blueprint solution was a very good example of how the case management capabilities in IBM Business Automation Workflow can accelerate development cycles, even for very complex projects. We only needed to develop a few custom widgets—the vast majority of the functionality we implemented uses standard features that come with the case management capability in IBM Business Automation Workflow out of the box.”
Samuel Myllykangas says: “Where Elinar really added value was in understanding how our geneticists do their job, and working with them to design a workflow that would let them be more productive and efficient.”
He adds: “Also, since we handle confidential medical data, it was imperative to ensure that the solution would protect patient privacy and comply with healthcare and pathology standards such as HIPAA, CLIA, CAP and ISO 15189. Elinar did a good job of leveraging IBM Business Automation Workflow’s robust security framework to help us meet these objectives.”
The case management capabilities in IBM Business Automation Workflow are integrated with the BigInsights solution via IBM Spectrum Scale™, which provides an extremely scalable file system. All the components run on bare-metal servers in SoftLayer®, IBM’s flexible hybrid cloud platform, and integrate with both the company’s on-premise sequencing systems and other cloud systems such as PubMed and similar online medical databases.
Samuel Myllykangas comments: “With our ambitious growth objectives, the ability to run the entire solution in a hybrid cloud environment was a huge advantage. SoftLayer gives us the ability to scale seamlessly as demand for our services increases—and as we expand into new markets, we are expecting to see a big increase in the number of samples we process over the coming months.
“Also, we’re a company of geneticists and clinical specialists, not IT people—we don’t want the burden of managing IT infrastructure. SoftLayer gives us the confidence that IBM’s infrastructure experts will keep our systems running smoothly while we focus on our core business: delivering fast, high-quality genetic testing services to our clients.”
Achieving business transformationWith the new solution in place, Blueprint has already been able to increase productivity by 30 to 50 percent, and expects to see a three-fold increase in overall processing speed and capacity as its analysts gains more experience.
“Over time, we expect to build up a knowledge base that will help accelerate the interpretation of more straightforward cases,” says Samuel Myllykangas. “This will not only speed up the interpretation process—it will also enable our analysts to spend more time focusing on the more unusual and interesting cases, where their expertise really counts.”
Since the solution also encourages a more structured and standardized approach to the interpretation and reporting process, it also makes it easier for Blueprint to train new analysts and ensure that everyone is complying with the appropriate lab standards and regulations.
“The process gives our analysts confidence that they have followed all the correct protocols and haven’t overlooked anything,” notes Samuel Myllykangas. “It also gives us complete traceability, so it is easy to audit our operations and demonstrate proper compliance.”
As a result of accelerating its interpretation and reporting processes, Blueprint has been able to scale up its product offering from 20 tests to 400 within just seven months.
“By eliminating one of the biggest operational bottlenecks, we have been able to achieve our goal of transforming our business from a cardiovascular specialist to an all-encompassing genetic testing service that can test for a comprehensive range of inherited disorders,” says Samuel Myllykangas. “We can now provide the one-stop shop that our clients have been looking for—giving us a significant competitive advantage.”
He concludes: “When we founded our business, it was common for genetic sequencing providers to take months or more to deliver results. As standard, we offer to deliver results within 21 days, and now that we have the IBM solution, we’re achieving average turnaround times of just 14 days.
“We see ourselves as raising the bar for the whole industry, and as we continue to build on the solution we’ve developed with IBM, Elinar and Enfo Rongo, we’re excited to see how much further we can raise standards in the future.”
About Blueprint Genetics
Blueprint Genetics is changing the field of clinical diagnostics by providing accessible and actionable genetic testing covering all medical specialties. The company seeks to support healthcare professionals around the world in providing the best care for patients and families with rare inherited diseases by delivering answers using accurate, fast, affordable and comprehensive genetic diagnostics.
- Case Manager
- Cloud Services - IBM Cloud infrastructure F2F (Cloud BU)
- IBM Spectrum Scale
Take the next step
Elinar is a specialist in IBM Enterprise Content Management solutions, which focuses on understanding its clients’ needs and developing efficient solutions to solve business problems. The company’s mission is to help its clients become stronger and more competitive through exploiting and processing their main resource: knowledge. To learn more about products, services and solutions from Elinar, please visit elinar.com. Founded in 2006, Enfo Rongo is a growing information management company operating in Finland, which specializes in helping organizations to manage data. It provides value-added solutions for data warehousing, big data, master data management and business intelligence, performance management and predictive analytics. To learn more about products, services and solutions from Enfo Rongo Oy, please visit rongo.fi. IBM Analytics offers one of the world's deepest and broadest analytics platform, domain and industry solutions that deliver new value to businesses, governments and individuals. For more information about how IBM Analytics helps to transform industries and professions with data, visit ibm.com/analytics. Follow us on Twitter at @IBMAnalytics, on our blog at ibmbigdatahub.com and join the conversation #IBMAnalytics.