Accelerating genomic research with High Performance Computing
High performance computing is providing researchers with processing power that was almost unthinkable in the recent past. The Digital Research Team is behind the University of Aberdeen’s supercomputing clusters, Maxwell, and Macleod, boosting the capability and capacity of both research and teaching across the university.
The team also supports the Centre for Genome Enabled Biology and Medicine (CGEBM) to deliver highly valued genomic services to staff, students, and external clients.
Modern science is generating huge amounts of data, requiring more and more computing power to process and analyse. The High Performance Computing Cluster (HPC) helps University of Aberdeen researchers keep pace, offering over a thousand desktop computers’ worth of power, and able to complete a year’s work of a single desktop computer in just one day.
“Instead of a computation taking weeks or months to do on single computer, it can potentially be done in hours or days,” says Michael Chung, Technical Digital Research Specialist in the Digital Research Team.
“And it’s not just that we have a large amount of resource. We can also have many projects running in parallel to make things even more time efficient. There could be 10s, if not 100s of separate projects running at any one time.”
Some researchers are experienced in using HPC and we can get them up and running quite quickly. Others may need a little more assistance and the team are happy to help.
“Once set up, they can take advantage of our rich resources to drive forward leading-edge research,” says Michael.
Genomics Analysis – Working in Partnership
One team that needs access to both the research and teaching aspects of the HPC is the Centre for Genome Enabled Biology and Medicine (CGEBM), which provides genomics analysis and related services and training across the university. The CGEBM has upwards of 100 programs installed in the HPC to support research.
“The High Performance Computing Cluster is essential for our work. The Digital Research Team really help us deliver our research, training, and teaching activities across the University and with external stakeholders.
“We work on anything related to genomics or bioinformatics, supporting a broad range of different types of research. The types of data that we generate – genomic data – are large, complex datasets. It’s simply not possible to do most of these types of analyses on a standard desktop because it doesn't have enough computing power.
“My team has specialist expertise in genomics data, what the programs are doing to the data, understanding the outputs and how the tools work.”
Genomics Teaching - developing researchers of the future
Another demand on the HPC is teaching, as big data projects become increasingly common across disciplines. To support skills development, the Macleod cluster is dedicated to teaching.
The CGEBM uses the Macleod cluster to train students, delivering Advanced bioinformatics and genomic sequencing courses to PhD and Masters’ students.
Through this course, attendees gain an understanding of how to use the technology in the context of genomics datasets and how it will benefit their future research.
“The Digital Research Team’s expertise in administration of the cluster and deployment of a wide range of bioinformatics software, with often complex dependencies, in a streamlined set-up to facilitate use by beginners and experts alike, provides invaluable support for our work. So we work together, often in an iterative process, to get things up and running,” says Elaina.
The Digital Research Team facilitates the delivery of genomic services, ensuring research, teaching requirements, and skills acquisition are well supported. We provide the technology (in-house managed hardware and software) and support (user management, software support, and troubleshooting), leaving the complex genomics bit to Elaina and her very capable team.