Genomics to hit mainstream with AI and $100 genome

The dramatic drop in the cost of genome sequencing, combined with rapidly evolving artificial intelligence, is moving precision medicine into mainstream healthcare.

Sanjay Chikarmane, senior vice president at Illumina, told an briefing at HIMSS17 in Orlando on Monday that the US$100 genome is now in sight.

Illumina is an US company focused on genetics sequencing and analysing big data for biological insights.

The company is also the major partner for the UK Government’s 100k Genome project, providing most of the infrastructure through a £78 million partnership with the Genomics England.

During the briefing, Illumina announced a new partnership with Philips. Illumina will use Philips new genomics AI platform to analyse genomics data, identify key mutations and provide the data into clinical workflows.

IBM’s Watson Health is also working on the AI genomics initiative with Illumina.

“Our mission is to improve human health through sequencing at a massive scale.  The first human genome sequenced ten years ago, took years and cost $3 billion,” said Chikarmane.  “We can sequence in less than a day and with our latest instrument it already costs less than $1K.  We are now on our way to the $100 genome.”

After the briefing, Chikarmane told Digital Health News that he could see the price of DNA sequencing falling below $100 in the future.

“We all know the potential is tremendous, barriers of time and cost have been overcome.”

“But to make it mainstream we have to be able to analyse the data to identify mutations that can be treated, and identify the most effective drugs that can be used to treat patients.  And this has to happen in the electronic medical record.”

The ability to analyse the tsunami of genomics data generated is now the greatest barrier to progress, he said.

To interpret the vast amounts of genomics data today requires sophisticated, highly trained bioinformatics specialists.

“It’s the clinical back end and interpretation that is the limiting factor and that’s where AI has such potential.”

“It can currently take 15 hours of a geneticist’s time to interpret one patient DNA sequence,” Chikarmane said.

“This is why it has been the realm of academics so far, using very highly trained geneticists, clearly this is not scalable… how do you bring it to community hospitals? How do you make mainstream? You have to use AI.”

Chikarmane said that genomics has to become embedded in clinical workflows if uptake was to move out of research labs.

“To be mainstream genomics has to be at the point of care and with radiology and pathology reports.”

Back in England, Genomics sequencing in the NHS has been supported through 100k Genome project. In 2014 the Government announced £300 million funding to support 11 Genome centres, which would be expected to sequence 10,000 genome by 2017.

That target was later pushed back to 2018.