Saturday, March 26, 2011

Where robots labour to overcome genetic disease

25 March 2011 by Peter Aldhous, Mountain View, California

I'M PEERING through an internal window into an eerie blue-lit room packed with high-tech machinery. The temperature inside is held at 28 °C and the humidity is high. Technicians scan computer screens to ensure the robots are happy, but seem oblivious to my presence.

Replace the blue light with green and it would resemble the interior of one of Star Trek's Borg cubes. Thankfully, the technicians look nothing like cyborgs, and the goal is not the assimilation of humanity. Rather, this is the production line that Complete Genomics, a start-up in Mountain View, California, bets will revolutionise the discovery of disease genes.

It is already the world's largest human genome sequencing factory. In a room half the size of a tennis court sit 16 robots that sequenced the genomes of 800 people last year. Going full tilt, they can now churn out 400 genomes a month.

The current price, offered to researchers and pharma companies but not yet to private individuals, is $9500 per genome; place an order for 1000 or more, and it drops to $5500. When you consider that the first human genome was completed a decade ago for billions, DNA sequencing has come a long way, fast.

As I survey the scene, Jennifer Turcotte, the company's marketing chief, explains why it looks different from other sequencing labs. DNA is usually read inside hermetically sealed machines, but here the robots work with their guts exposed, for ease of maintenance. The heat and humidity suit the biochemistry of the sequencing reactions, and the dim blue light avoids frequencies that would bleach the fluorescent probes used to detect each letter of the genetic code.

The technicians wear clean-room gear, as dust would interfere with reading the sequences. Unless something goes awry, there is no need for them to intervene. The robots add the required reagents, and manoeuvre the samples so that a camera can record the light signals that reveal the DNA sequence of 70 bases at a time.

The formidable computation needed to assemble these snippets into 3-billion base-pair human genomes is done at a fully automated data centre about 20 minutes' drive away in Santa Clara - electricity is cheaper there, and data storage is charged by the kilowatt-hour, explains Clifford Reid, the company's CEO.

With the production line essentially running itself, most of the 185 staff are busy improving the company's sequencing technology, or liaising with customers. "We need people to interact with people, but not to interact with the DNA," says Reid.

In a cute twist, the company has even automated its reception area. When I arrived, I was greeted by a computer terminal, which asked for my name and who I had come to see. A label printer spat out a visitor badge while an email summoned Turcotte to lead me into the inner sanctum.

The culture of automation has a serious scientific goal. Geneticists had hoped that mutations determining our susceptibility to disease would emerge from limited scans, which record common variants at some 1 million positions across the genome. But the discoveries so far explain a small part of the heritability of many conditions. Gene hunters are starting to hit a wall.

Looking for the missing mutations means sequencing entire genomes and pinpointing rare anomalies that are inherited with the disease in question. The principle was exhibited last year by a team led by Leroy Hood of the Institute for Systems Biology in Seattle, which narrowed to a list of four the mutations responsible for the craniofacial condition Miller syndrome in one affected family (Science, DOI: 10.1126/science.1186802). Complete Genomics did the sequencing.

Several companies are pushing the envelope of cost and speed in DNA sequencing. But Complete Genomics is unusual in tailoring its technology to the task of churning out whole human genomes, and deciding not to sell machines but to offer a contract sequencing service. Hood sees little point in scientists doing the work: "We want to put our efforts into developing the tools to interpret the information."

Hood sits on Complete Genomics' scientific advisory board, but the company is also winning customers who don't have a personal interest in its success. They include the US National Cancer Institute, which is comparing the sequences of normal and cancerous cells from the same individuals to seek mutations behind childhood cancer.

For disease-causing mutations, will resistance be futile?

Source and/or read more:

Publisher and/or Author and/or Managing Editor:__Andres Agostini ─ @Futuretronium at Twitter! Futuretronium Book at