[tt] Telegraph: Why google wants your genes

[Premise Checker]

Why google wants your genes
http://www.telegraph.co.uk/connected/main.jhtml?xml=/connected/2007/10/06/dlgoogle06.xml
[Thanks to Sarah for this.]
7.10.6

DNA fingerprinting could turn the titan of web-searching into a
medical behemoth, says Emma Hartley


As if gauging the nation's receptiveness to new technology, Lord
Justice Sedley suggested recently that the UK's whole population
and its visitors should have their DNA added to a Home Office
database that already holds genetic information about four million
people - five per cent of the UK population, and the highest
proportion of any state in the world.

Sedley is known for his progressive views and has a record on the
bench of upholding civil liberties, so this was electrifying stuff.
Not only would the measure confer obvious advantages on the police,
while getting around the objection by civil libertarians that
ethnic minorities are disproportionately represented on the
database, it also promised a practical use for a technology so new
that the four million sets of data were collected before most of us
even knew it was happening.

Ever since the completion of the Human Genome Project in 2003, in
which the first whole set of 23 human chromosomes was decoded into
its constituent bases, scientists and biotech businesses have been
agog at the possibilities. Spotting an opportunity, a group of new
companies has begun offering to "mine" your genes for information
about your ancestors.

One is Oxford Ancestors (oxfordancestors.com), started by Prof
Brian Sykes of Oxford University, which will tell you from which of
36 geographically located "tribes" your ancestors originated, all
for £180. Cambridge University offers a similar service for £30
less.

But the decoding of the human genome promises much more - just ask
a geneticist. An entire history of life on Earth is buried within
the cells of our bodies if you possess the skill to interpret it,
as is information about our collective longevity, degeneration and
ultimate demise. The big question, though, is "What does it mean
for me?" And are we prepared to pay to find out?

Ever alert to the next big thing, Google recently bought a $3.9
million (£2 million) minority stake in a Californian company called
23andme (23andme.com), among whose founders was Anne Wojcicki, who
married Sergey Brin - one of Google's co-founding geeks - earlier
this year.

This marriage is the stuff of Silicon Valley legend, since Ms
Wojcicki's sister rented out her garage to the "Google guys" when
they left Stanford University to pursue their start-up. But this is
not a story about nepotism.

What 23andme will apparently be offering when it launches later
this year is a personalised service that will let individuals have
their DNA read and the information stored, so that when medical and
theoretical advances relevant to them are achieved, they will be
notified. Many people think of Google simply as a really fast
search engine, but its stated mission has long been to organise all
the information in the world: see David A Vise's book The Google
Story (2005) for more on this.

So its interest in the human genome is a natural extension of its
original goal, with the extra spice that computing and genetics are
both industries growing exponentially. When you take into
consideration Google's unparalleled server power, the tie-up
between the two companies could represent a historic breakthrough
in personalised medicine.

Information about one's susceptibility to specific diseases,
vulnerability to the side-effects of existing medicines and
suitability for new ones coming on to the market could all be
gleaned from this kind of genetic mapping.

In June, for instance, a paper was published by the Wellcome Trust
confirming that they had identified the genetic pathways of seven
human diseases: bipolar disorder, coronary artery disease,
hypertension, rheumatoid arthritis, Crohn's disease and Type 1 and
2 diabetes. This opened up the field for more accurate treatments.

British companies are directly involved in this burgeoning
industry. Researching this article, I was advised that one of the
best people to talk to in the UK would be Dr Shankar
Balasubramanian.

Yet trying to get hold of him, the apologetic message was returned
that he would be unable to speak to me because a company he had
founded in Cambridge, called Solexa, was bought out by an American
company called Illumina for $600 million in November last year and,
since Illumina (and therefore he) was involved in the 23andme deal,
his hands were tied until the launch.

Dr Richard Durbin of the Welcome Foundation's Sanger Institute in
Cambridge was on Solexa's scientific advisory board and explained
that until now, the main obstacle to personalised medicine has been
money.

"Each person's DNA is 3,000 million bases long and decoding the
whole lot costs around $10 million. But the technology Solexa was
developing is known as high speed genome sequencing and it's
bringing the cost down very considerably. There are currently three
companies using it that I'm aware of: Illumina, Roche and Applied
Biosystems. But in reality there is no commercial demand for
sequencing an entire genome.

"There are only 12 million positions on our DNA that differ from
each other - the rest are the same in everyone. It is these 12
million that make us individuals and decoding this, in a process
called genotyping, currently costs about $1,000. The technology is
developing so fast that very soon it will drop to the low hundreds
of dollars."

The Home Office's DNA fingerprint uses only 10 discreet areas of a
chromosome, costs less than £100 per print and has a theoretical
chance of one in a billion of being wrong, although this was
undermined by recent Home Office confirmation that around 550,000
false or wrongly spelled names had been submitted with DNA samples
by the police, making them of little practical use in court.

There are ethical considerations, of course. Genes can be switched
on and off, so telling someone that they have a predisposition to
cancer, for instance, would surely have to be accompanied by advice
on how to minimise the risk of developing it.

There is also the issue of what happens to the stored DNA data,
since any company holding such information (the decoded DNA and the
identity of the person involved, to discover how those genes are
expressing themselves) would possess a goldmine of information for
the pharmaceutical industry.

The cost and availability of treatments would be a further ethical
sting in the tail. But it is hard to resist the thought that such a
service will find a mass market. Can you imagine the conversation
that would ensue at a dinner party of genetically decoded
hypochondriacs?