Google works with the Garvan Institute for faster sequencing for 14,000 genomes

More accessible and cheaper gene-based diagnosis for millions of Australians closer to reality, as Google partners with the Garvan Institute.

Garvan institute of Medical Research Associate Professor Sarah Kummerfeld. Picture: Supplied.

Google and the Garvan Institute have collaborated to develop radically faster genome sequencing, bringing more accessible and cheaper gene-based diagnosis to millions of Australians.

Associate professor Sarah Kummerfeld, from Australia’s Garvan Institute of Medical Research, said it took 20 years and cost $3bn to sequence the first genome.

However, the collaboration between the institute and Google led to 14,000 genomes being processed, or understood, in less than two weeks. This was unprecedented in Australia, she said.

Google has provided the Institute with cloud infrastructure and links to computers that can process the genome faster.

Dr Kummerfeld said the new processing speed was a massive achievement given the genetic code people inherited from their parents was about six billion letters long.

She said Garvan’s high performance infrastructure would have taken more than a year to process those 14,000 genomes, however storing the genomes in the cloud meant massively fast supercomputers in other countries could access them and process them.

She said research could be shared with other qualified researchers through the use of the cloud, provided by Google.

Dr Kummerfeld said people could take more targeted preventive measures and potentially have their illnesses diagnosed more readily when doctors knew the diseases they were vulnerable to, using information derived from genome sequencing.

She said that by sequencing a person’s genome, you might establish they were at risk of cancer and recommend additional screening such as mammograms. “It means you can catch these diseases early.”

She said Garvan’s charter was to use information encoded in the genome to prevent, diagnose and treat diseases. Deciphering the disease a person might contract involved comparing their genome with a “reference genome” of a healthy person.

“In the case of cystic fibrosis, you can have one gene with one letter change and then you get disease, but things like cancer and heart disease involve many, many changes across the genome that increase your risk.”

She said the processing of the 14,000 genomes was a proof of concept.

Instead of costing $3bn as the first sequencing did, the cost is decreasing to under $1000. “If you include the interpretation of the genome, we’re looking at around $2500,” she said.

“Because costs have come down so much, people are turning to the genome first.

“Rather than going through dozens and dozens of tests that might take years and there‘s no treatment in the meantime, they do the genome sequence first and they can get a diagnosis much faster.”

Dr Kummerfeld said the data set generated from the project would be used by researchers at the Centre for Population Genomics, a partnership between Garvan and the Murdoch Children’s Research Institute.

Alister Dias, vice president Google Cloud, ANZ, said open-source collaboration was “core to Google Cloud’s DNA”.

“We know that solving some of the world’s biggest challenges doesn’t happen in a vacuum,” he said.

“Google Cloud’s scalable and secure infrastructure enabled Garvan to analyse massive amounts of biomedical information at unprecedented speed. The potential of this research to quickly understand and find cures for gene-related diseases is significant, and one we’re proud to be a part of.”