7
$\begingroup$

In the price curve of the genome, from NIH:

price curve

NIH explained the abrupt drop at 2008 thusly:

beginning in January 2008... the sequencing centers transitioned from Sanger-based (dideoxy chain termination sequencing) to 'second generation' (or 'next-generation') DNA sequencing technologies.

The NIH did not give any details as to why there was a similarly precipitous drop in 2015, or why there was such a saddening (for me!) plateau after that. Please explain them to me?

$\endgroup$

3 Answers 3

6
$\begingroup$

This graph from the Broad's Opinionome blog (ugh) is somewhat more annotated:

Annotated graph with sequencing costs dropping in 2007 and 2014

As noted elsewhere, the precipitous drop in 2007 is almost certainly due to maturing next-gen sequencing (NGS), in particular Illumina. Illumina acquired Solexa in 2007, which offered gigabase-level sequencing ability. In 2014, President of Illumina Francis de Souza essentially said as much:

During the EmTech conference, De Souza said Illumina’s success was due to a “hard pivot” the company made in 2006, when it got into the DNA sequencing business by acquiring Solexa, a U.K. startup, and bet its fortunes “on a technology with no sales, that no one knew if it would work.”

Essentially, by 2007/2008 you had stellar technology+company hard-selling it = adoption

I'm less certain about the less-dramatic drop in 2014/2015, but I agree with Franck that it's likely largely be due to the introduction of the Illumina HiSeq X Ten in early 2014. You can read some corporate spam, but it was indeed a big deal. There were other technologies that probably played a role (new PacBio system released in 2013, etc.) but Illumina was and is top-dog.

None of this takes into consideration increased interest in such research, or the faster computers, larger SSDs, and expansion/normalization of multi-core systems that makes working with such data easier.

$\endgroup$
4
  • 2
    $\begingroup$ The "opinionome" is the oddest use of "-ome" I've seen so far lol $\endgroup$ Commented Jul 5, 2019 at 11:30
  • $\begingroup$ @MaudPieTheRocktorate This is likely intentionally tongue-in-cheek, but a friend and I used to keep a list of wild -omics words in published papers, and... yeah, there are some categorically absurd ones. $\endgroup$
    – Amory
    Commented Jul 5, 2019 at 13:44
  • $\begingroup$ Would you give a few examples? For the lolz of posterity. $\endgroup$ Commented Jul 9, 2019 at 10:17
  • $\begingroup$ Some I consider begrudgingly fine — injectisome, translatome, inflammasome — while others are bad and feel like they're trying to make "fetch" happen — promoterome, ligandome, connectome — and others are just beyond — HAR-ome, imagenomics, screenome. $\endgroup$
    – Amory
    Commented Jun 18, 2020 at 9:53
4
$\begingroup$

The graph is in contradiction with the data in the Excel spreadsheet the same webpage provides (https://www.genome.gov/sites/default/files/media/files/2019-06/Sequencing_Cost_Data_Table_Feb2019.xls (mirror)), which I plotted below:

enter image description here

enter image description here

Update on 2019-11-17: I contacted [email protected] and the graph on https://www.genome.gov/about-genomics/fact-sheets/Sequencing-Human-Genome-cost is now fixed:

enter image description here

https://en.wikipedia.org/w/index.php?title=\$1,000_genome&oldid=902062326#Commercial_efforts to some of the dates on the reaching the 1000 USD/genome goal, and doesn't indicate any plateau:

In January 2014, Illumina launched its HiSeq X Ten Sequencer, claiming to have produced the first \$1,000 genome at 30× coverage. Some researchers hailed the HiSeq X Ten's release as a milestone – Michael Schatz of Cold Spring Harbor Laboratory said that "it is a major human accomplishment on par with the development of the telescope or the microprocessor". However, critics pointed out that the \$10 million upfront investment required to purchase the system would deter customers. Furthermore, the \$1,000 genome cost calculation left out overheads, such as the cost of powering the machine.[25] In September 2015, Veritas Genetics (co-founded by George Church) announced \$1,000 full-genome sequencing including interpretation for participants in the Personal Genome Project.[26]

In April 2017, the newly formed European company Dante Labs started offering the WGS for \$900.[27][28] In 2017, Beijing Genomics Institute began offering WGS for \$600.[29] In July 2018, on Amazon Prime Day, Dante Labs offered it for \$349.[30] In November 2018, around the time of Black Friday, Dante Labs offered WGS for the first time less than \$200,[31][32] and Veritas Genetics for two days for the same price of \$199 offered WGS limited to a thousand customers.[33] In March of the same year, geneticist Matthew Hurles of Wellcome Sanger Institute noted that the private companies, including Illumina,[34] are currently competing to reach a new target for WGS of only \$100.[35]

Page summarizing current prices: http://arep.med.harvard.edu/gmc/genome_services.html (mirror).

Update (2019-07-03) https://www.cnbc.com/2019/07/01/for-600-veritas-genetics-sequences-6point4-billion-letters-of-your-dna.html (mirror):

Veritas is lowering the cost of a full genome sequence from \$999 to \$599 and believes within two years it will be pricing in the \$100–\$200 range.

Another price progression plot from https://youtu.be/jH87sfVD36M?t=554 ("George Church - Cognition Genes. Aging Reversal. Lab-built Brain Components"):

enter image description here

$\endgroup$
3
  • $\begingroup$ Illumina launched HiSeq 2000 in early 2010 after purchasing it from Solexa in 2007 dx.doi.org/10.1155/2012/251364 $\endgroup$
    – Dexter
    Commented Jun 29, 2019 at 5:27
  • 1
    $\begingroup$ By "contradiction" do you mean the mistake in the y-axis labeling? $\endgroup$
    – tyersome
    Commented Jun 29, 2019 at 18:28
  • $\begingroup$ @tyersome yes that seems to be the issue $\endgroup$ Commented Jun 29, 2019 at 20:29
1
$\begingroup$

I'm not 100% sure, but I think Illumina launched their HiSeq machine around that time, massively increasing the sequencing output in one run. That dip might be related to that.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .