April 07, 2004
DNA Sequencing Costs Continue to Decline

Biotech instrumentation company Affymetrix has announced a new instrument for mapping DNA Single Nucleotide Polymorphisms (SNPs) which Affymetrix claims is much cheaper to operate than previous instruments designed for this purpose. SNPs are locations in the genome where not all humans or all members of another species have the same the DNA letter. The new Affymetrix instrument will reportedly lower the cost of human DNA SNP testing to 1 penny per DNA letter.

The 100K builds on the innovative, scalable, easy to use assay that Affymetrix pioneered with the GeneChip Mapping 10K Array. The 100K allows researchers to genotype over 100,000 SNPs using just two reactions. Previously, genotyping 100,000 SNPs would have required 100,000 PCR reactions, a hurdle that made this kind of research impractical. Before the advent of 100K, the commercial product for genotyping the most SNPs was Affymetrix' Mapping 10K.

“The power of 100,000 SNPs in a single experiment is enabling researchers to attempt unprecedented genetic studies at a genome-wide scale,” said Greg Yap, Sr. Marketing Director, DNA Analysis. “The GeneChip Mapping 100K Set is the first in a family of products that will enable scientists to identify genes associated with disease or drug response across the whole genome instead of just studying previously known SNPs or genes, and to study complex real-world populations instead of simple ones. To do this, we are making large-scale SNP genotyping not only quick and easy, but also affordable -- about 1 cent per SNP.”

About half of the SNPs on the 100K set are from public databases, while the other half are from the SNP database discovered by Perlegen Sciences, Inc. All of the SNPs on the 100K set are freely available and have been released into the public domain. Because the assays and arrays used in the 100K set are extremely scalable, more SNPs from both public sources and the Perlegen database will be added to next generation arrays.

Just a couple of years ago the cost of SNP assays was at about 50 cents per SNP. Industry analysts at the time were predicting SNP costs to fall to 1 cent per SNP within 2 years and sure enough if Affymetrix's press release is realistic those costs are just about to fall to the predicted 1 cent per SNP.

To put that number in some useful perspective, there may be about 10 million SNPs in humans but perhaps about only a half million SNPs that are in areas of the genome that affect human function. Most of the SNPs are in junk regions. The 500,000 estimate of functionally significant SNPs is a scientific guess at this point and could easily be off by a few hundred thousand plus or minus. But once those 500,000 SNPs are identified (which could easily take a few years yet) the cost of 1 penny per SNP to test them all would cost about $5,000.00 per person in US dollars to test a single person's complete genome for SNPs. The real cost would be higher since samples of portions of the genome would need to be isolated for experimental runs. The real cost might easily be several times that amount. That still wouldn't provide a complete picture of a person's genome since there are few other kinds of genetic variations (e.g. short tandem repeats or STRs). But SNPs are responsible for most genetic differences within a single species.

At this point the decline in SNP testing prices is useful chiefly to scientists since we don't know which SNPs are important let alone how they are important. Still, the lower costs for SNP testing will accelerate the rate at which important SNPs are identified and that will bring closer the day when it makes sense to for individuals to go get complete SNP testing.

Full DNA sequencing is also at about 1 cent per DNA letter.

Our price is competitive at just $.01 per base per sample, all inclusive. For example, total sequencing for a 7 Kb region on 48 samples would cost a total of 7000 x 48 x $.01 = $3,360. This pricing structure makes SNP discovery project estimates straight forward, avoiding the possible uncertainty of estimating the exact number of PCR amplicons, sequencing reactions, etc.

This price would make a complete 2.9 billion letter human genome sequencing for a single person cost about $29 million dollars. But it would really cost several times more than that since sequencing has to be done multiple times to catch errors in the sequencing and there are sections of the genome that are hard to sequence. Since whole genome sequencing is much more epensive ways to lower the cost of the SNP testing is what is attracting the most interest. When SNP testing costs by fall another order of magnitude the cost will be low enough for at least the more affluent among us to want to pay for personal SNP tests. Once the costs fall yet another order of magnitude SNP testing for the masses will become common place. My guess is that given the rate at which SNP testing costs continue to fall and the predictions of industry figures of much cheaper SNP testing we are at most 10 years away from the point of general population large scale SNP testing..

Note: The Affymetrix press release link was found in a post on the Personal Genome blog.

Share |      Randall Parker, 2004 April 07 02:53 AM  Biotech Advance Rates

Bill Bailey said at May 19, 2004 11:03 AM:

I have been identified as NRY DNA Haplogroup Q.....Can this chip test identify SNP's down to the
YCC figure haplotype level?

Miriam Hughes said at February 8, 2005 6:54 PM:

Our Rose family group has been spending much time & effort on trying to find our Native American ancestor. We can follow our written Rose history to Marshfield Plymouth Colony 1653 to Joseph Rose who was married in that year to daughter of Edward Bumpus who came on 2nd ship. We have 10 DNA tests that indicate we are of common ancestor. I feel that perhaps we are going about this wrong and that our ancestor did come to this land as an Englishman. Family Tree DNA now has 1/2 of us as Q since they did the SNP test. I have submitted my cousin's test for the SNP. I expect that it will be classified as Q not the Q3 as it is now. This Joseph Rose's son married Rebecca Burge(Burgess) daughter of Thomas who was governor of the Colony. I just find that would make a number of revisions to the history of our Pilgrim ancestors customs as we beleived them to be.

Some of the websites on DNA seem to find Q y-DNA in quite a few places including Scotland where other Roses are known to come from. I would love to have some input from those who know so much more than I. Thank You, Miriam Rose Hughes

Dave Rowlandson said at October 22, 2005 5:47 AM:

Have a look at the National Geographic web site under The Genographic Project

Post a comment
Name (not anon or anonymous):
Email Address:
Remember info?

Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright ©