IBM LSF paves the way for NHS fast mapping of 100,000 cancer patient genomes

"Nobody else has the size or scale" says Genomics England

IBM's quest to develop systems to swiftly map the genomes of the UK's 100,000 cancer sufferers continues apace, and Computing caught up with both the vendor and customer Genomics to find out more about this crucial project.

Speaking to us ahead of sessions at Computing's IT Leaders Forum Learning from the Leaders in Big Data on 23 June 2016 in London, head of informatics infrastructure at Genomics England, Dave Brown, identified the need for "high volume" genetic sequencing.

"The project is to create 100,000 raw human genomes of NHS patients suffering from cancer," said Brown.

"The aim of that is to pave the way for progress in the National Health Service for genetic medicine, but also to look at how we can start to build a technology platform that can allow the sequencing at high volume."

At the moment, Brown explained, sequencing all cancer patients in the UK would take "a minimum of 20-30 days, if you're lucky".

To be viable for therapeutic benefit, this would need to be cut to 10 days, minimum, he continued.

The answer lies in employing a better IT infrastructure, and this is where IBM's workload management platform, LSF, comes in.

"It allows us to maximise the throughput on our computer system," said Brown.

Janis Landry-Lane, worldwide software-defined life sciences industry lead at IBM, said the question IBM and Genomics now need to tackle together is whether we can "design a computing architecture that not only supports today's workload environment, but all of these workloads that we see on the horizon".

"We see a lot of changes coming, and we want to make sure that LSF is absolutely compatible with anything that might happen in the future," said Landry-Lane.

"New platforms and paradigms are going to arrive as new algorithms are written, and this works wider in human biology than just Genomics' use case.

"We at IBM are not doing it all," she said, explaining that pan-omics information management firm Databiology - a Genomics and IBM partner - is another firm working on crucial medical big data.

The scalable reference architecture LSF offers supports on-premise and cloud, offering a flexibility medical customers need depending on the task at hand. The need for heavy lifting can vary depending on task.

Additionally, the AI-driven Watson Health Cloud is another service that can be applied to health work.

"We're talking to IBM at the moment about evaluating the Watson technology for clinical interpretation and decision support," said Brown.

"It's one of the few interpretation technologies that actually integrate it with clinical decision support. Most of the stuff in the bio and genetics area is very biased towards genetics, but Watson brings clinical technology in," he said.

"The other thing is, it's paving the way for rolling out clinical genetics across the National Health Service, and you need partners of the scale and size who could actually roll a solution out across the whole NHS, for 65 million people.

"Most of the other people in this area are VC-funded, 50-people companies primarily based in the US, with no UK or Europe-based ability to deploy and support."

To hear more from both Dave Brown and Janis Landry-Lane, as well as discover further insights from key Computing big data research, book your place at Learning from the Leaders in Big Data on 23 June 2016 in London. We look forward to seeing you there.