Microsoft rolls out genomics service to help doctors identify treatment

Tech giant Microsoft Corp.'s healthcare arm has used artificial intelligence and cloud computing to come up with a new service that will, among others, help medical workers choose effective treatment for diseases such as cancer based on a massive repository of genomic data.

Called Microsoft Genomics, the service is a direct result of a 2016 diktat by chief executive Satya Nadella to infuse new-age tech into healthcare.

Bob Davidson, a principal software architect in Microsoft’s genomics group, explained the service: "By analysing genomic data from a patient’s healthy tissues as well as tumour tissues, a physician will be able to select the treatment that will be most effective based on comparison with data from other cancer patients, including treatments and outcomes."

The base research for the service, however, had already begun in 2010, outside of Microsoft, when computational biologist Jinghui Zhang and her team at St. Jude Children’s Research Hospital in Memphis started uploading anonymised genomes of their patients’ healthy and cancerous cells to public data repositories. This was done so that scientists and researchers across the world could study the data in their search for cures to paediatric cancers.

However, people found it hard to download and use the data, given its sheer volume, Zhang said. “So, St. Jude started to seriously explore other ways to ease data sharing with the global research community,” she added.

That led to a collaboration with members of the genomics group in Microsoft’s research organisation. Zhang’s team provided St. Jude genomic data, which Microsoft put through a cloud-based pipeline in a technique known as ‘alignment and variant calling’. Genetic variants are what make individuals unique, from physical attributes to disease susceptibility.

So far, the collaborators have processed a half petabyte of genomic data and stored it on Microsoft’s cloud computing platform Azure for analysis. To give an idea, a half petabyte of data would fill 750,000 standard CD-ROM discs.

The St. Jude genomic data analysed through the pipeline and stored in the Azure cloud is the foundation for data-sharing platform Microsoft Genomics.