How low can a genome go?

The first computers ever made cost the equivalent of the GDP of a small country, and yet today similar processing power can be found in giveaway items from fast-food restaurants.

The first cars were for the super-rich only, as were telephones and radiograms and even healthcare.

Today all of these things are available to just about everybody: things get cheaper, and we expect everything to follow this path.

But what about the genome? It has long been a dream of the medical profession to have instant access to an individual's complete genetic profile, and (significantly) the knowledge to be able to interpret what that information means.

The concept of 'personalised medicine' is built largely around this notion, and where once this may have seemed fanciful, today it seems within reach.

Of course, medicine can be 'personalised' without recourse to a complete genotype.

The detailed analysis of certain key sequences on a specific chromosome of an individual is today a common enough procedure - not exactly routine, but far from unknown.

And, it might be argued, there is little specific need to sequence an entire genome, given that 99% of this is common to all humans.

Simple arithmetic shows, then, that the US$3billion cost of the Human Genome Project - the output of which was one more-or-less complete map of human being, could be reduced to one-hundredth of that by merely not bothering to resequence the universal parts of human DNA.

This would bring the cost of an individual genome to a mere $30million.

In fact, commercial genome sequencing can already be performed for a fraction of that.

According to GATC Biotech, the cost today is only $5million.

The company has recently launched a commercial genotyping service, and sees great potential for yet faster and cheaper services.

Last year, I interviewed a development scientist at Applied Biosystems about the potential for personalised medicine.

Surrounded by an impressive array of high-throughput machinery and laboratory automation technology, he offered a vision of the universal availability of a personal genome: individuals would send a sample to a laboratory, and the next day would receive a dataset of their own complete genome.

The data would probably be held on a remote database, but a personal back-up copy on a unique data DVD would be mailed.

Conveniently, this has just the right capacity to record a human genome.

The vision then, in early 2006, was for the 'one-day, thousand-dollar' genotyping service.

Since then, development seems to have accelerated and that target now looks less ambitious than it did a mere 18 months ago.

Technologies like the HumanLinkage-12 Genotyping BeadChip from Illumina promise the highest genotyping data quality at the lowest price, and clearly this impressive system is not the end of the developments that can be expected.

So, what is the end - and when can we expect it? GATC Biotech has this week set out its vision, reinforcing the view that costs will spiral downwards.

It promises to offer a complete human genotyping service, within the next decade, for just 500euro.

At today's exchange rate that equates to about $750, although the figures are perhaps skewed by volatile foreign exchange markets, and it seems reasonable to see this as the first pointer to the $500 genome.

One single human genome cost $3billion just four years ago, and will cost only $500 in ten years time.

This amazing deflation brings the cost to two hundred-thousands of one percent of the original.

But it is not a realistic comparison, when the first was clearly a prototype development.

Taking today's figure of $5million, and the projected $500 cost in a decade, gives a ten-thousand-fold improvement in cost over current technology.

And that makes other technology sectors, like electronics, look rather pedestrian.