Wanted: protein processors
Researchers need more muscular computers to pour through mounds of data whether it comes from new trials or databases of historical records.
"We desperately need a lot more computing power--that's a daily issue for us," said Quertermous of Stanford.
"We are measuring an entire genome sequence across thousands of individuals," he said. "You could use a thousand node computer cluster for months and not get an answer to the gene-by-gene interactions we are seeking," he added.
Specialized hardware could help researchers sort through as many as 300,000 known proteins to find still unknown variants that in small concentrations may have significant impacts on disease.
Today "there is no convenient way of characterizing all the proteins in a sample of blood, especially ones in low concentrations," said Avi Kulkarni, an advisor with Booz & Co. who moderated the panel.
"We need a technology platform that can allow us to identify hundreds and thousands of low [concentration] proteins on a purely discovery basis," said Kulkarni. "We will be able to use this knowledge to see disease patterns that we can only imagine today and, some of these proteins will be therapeutic," he said.
Currently researchers use techniques such as mass spectrometry to find the molecular weight of proteins and X-ray crystallography to describe their structure. "This is a very tedious process and it takes in the order of months to identify and characterize a new protein," Kulkarni said.
Long term, experts foresee a new class of wearable or portable consumer devices that monitor both proteins in the blood and a patient's environment to find patterns that will be key to treating them effectively.
"If patients can check biomarkers to see if a new diet is taking them in a right direction, that's a big opportunity," said Alex Bangs, chief technology officer of Entelos (Foster City, Calif.) which has developed computer modeling tools for genetic research.
"The answer is not just in the genes, but more likely in gene/environment interactions," said Quertermous.
In the short term, an emerging class of direct-to-consumer genetics companies is struggling to gain traction.
In June, the nascent sector had its first industry conference. However two companies seen as leaders in the field could not attend a panel on the topic here, at least one of them in the wake of bankruptcy.
"There's a big issue around how we make this a business and unfortunately we have two empty chairs highlighting this question," said Kristin Pothier, a vice president of consulting firm Health Advances (Weston, Mass.) who organized the panel.