Gene Synthesis Cost Update

While at iGEM this past weekend, I learned that GeneArt is now charging $.55 per base for ~1 kB synthesis jobs, with delivery within 10 days.

Here is an interesting tidbit: They only charged iGEM teams $.20 per base.  Anybody have any idea whether this represents their internal cost, and how much margin this might include?

Here is an updated plot for synthesis and sequencing cost.  No new data, just a new rendering.

(Update: 12 November, 2008.  There is a news piece in last week's Nature that claims Illumina's Genome Analyzer (GA1) was just used to sequence a whole genome in 8 weeks for $250K.  However, the paper describing that sequencing efforts says:

We generated 135 Gb of sequence (4 billion paired 35-base reads) over a period of 8 weeks (December 2007 to January 2008) on six GA1 instruments averaging 3.3 Gb per production run. The approximate consumables cost (based on full list price of reagents) was $250,000.

Thus the price does not include labor, and is not a true commercial cost (labor is only truly free for professors).

I am therefore not sure if/how this price can be compared to the prices in the figure below.

Update 2: I fixed the significant figure issue with the cost axis.  Alas, Open Office does not give great control over the appearance of the digits.)

carlson_cost_per_base_nov_08.jpg

Amyris Opens Biodiesel Pilot Plant

Amyris appears to be making good progress towards meeting their goal of getting biofuels to market by 2010.  They just opened their first pilot plant in California, with the aim of importing fuel into the US from Brazil within two years.  The output of the pilot plant will be used to gain EPA certification.  The announcement pretty well tracks with my previous posts about biofuels.

Here are a few graphs from the press release:

Amyris' diesel is characterized as a No Compromise™ fuel because it is designed to be a scalable, low‐cost renewable fuel with performance attributes that equal or exceed those of petroleum‐sourced fuels and currently available biofuels. Other attributes include:
  • Superior environmental performance: Preliminary analyses show that Amyris diesel fuel has virtually no sulfur and significantly reduced NOx, particulate, carbon monoxide and hydrocarbon exhaust emissions relative to petroleum‐sourced diesel fuel.
  • High blending rates: Because Amyris renewable diesel contains many of the properties of petroleum diesel, Amyris can blend the fuel at high levels ‐‐ up to 50  pecent ‐‐ compared with 10‐20 percent for conventional biodiesel and ethanol.
  • Compatibility with  existing infrastructure: Unlike many commercially available biofuels, Amyris expects to distribute its renewable diesel through the existing fuel distribution and storage infrastructure, thus speeding time to market while minimizing costs.
  • Adaptive: Amyris can produce its fuels from a broad range of feedstock including sugarcane and cellulosic biomass. It is starting with Brazilian sugar cane because it provides the most environmentally sound, economical, and scalable source of energy available today.

"This new diesel fuel has all the characteristics to make an important contribution toward solving our global transportation energy and climate crisis," said John Melo, chief executive officer of Amyris. "The opening of our pilot plant is a significant business marker for us, taking us one step closer to bringing our diesel fuel to market."

Craig Rubens at earth2tech provides interesting coverage, and his story notes:

Melo described the company's business model as "a capital-light model to scale up fast." The company plans to partner with existing ethanol plants and convert a portion of those partners' production capacity to make diesel and other chemicals using Amyris IP. The startup will then buy the products back from the refiner and take them to market, Melo said. The startup has already formed a joint venture with Santelisa Vale, Brazil's second largest sugar grower, called Crystalsev, which aims to produce 200 million gallons of fuel a year by 2011 at several of its existing ethanol plants at a price of less than $2 a gallon.

The Brazilian partnership, Melo explained, gives Amyris access to ports and ships to export the fuel. Amyris plans to import it to the U.S. and sell its to large customers, like Wal-Mart and the U.S. government. Foreign ethanol is hit with a 54-cent-per-gallon tariff as it comes into the U.S., but Amyris would be importing hydrocarbons, not ethanol, and therefore avoid the tariff. Amyris is already marketing other companies' biofuels in the Southeast to make sure its distribution channels will work.

To date, Amyris' strategy hasn't seamed particularly "capital light." The company has raised more than $120 million in capital (see previous coverage here and here) from heavy-hitting cleantech and biotech investors, including Kleiner Perkins, Khosla Ventures, TPG Biotech and DAG Ventures.

I understand the present need for scale, both physical and financial, and earth2tech's skepticism seems a bit naive.  Amyris is facing enormous competition, both from established petroleum companies and from other start-ups.  As I would not expect any of these companies to have a firm lock on IP surrounding biological production of fuels, Amyris must establish itself and its brand quickly and rely on first-mover advantage. (I wonder how thoroughly they are scrubbing the waste stream?  Dumpster diving for competitive intelligence takes on a new meaning here.)  Shell is dropping seven billion on upgrading a single refinery in Texas.  Amyris seems pretty light in comparison.

Writing at Cleantech, Emma Ritch provides an excellent tidbit: "The company has shelved its plans for a bio-gasoline".  "We're focused on the products with the highest value," Melo said. "We're not investing our resources in developing a bio-gasoline because we see the U.S. as the last gasoline-based economy."  That is particularly fascinating, as Melo is the former President of BP Fuels.  It is also a change since I heard Zach Serber speak at SB 4.0 last month in Hong Kong.  The fluctuating price of oil may be important here.

Unfortunately, Ritch mischaracterises the competitive landscape a bit: "Amyris plans to use the cheapest nonfood feedstock available, which for now means sugarcane... The company could also use algae for its biodiesel--much like Solazyme, LiveFuels, GreenFuel Technologies and many others."  In contrast to Amyris, the latter three companies are directly producting fuel in algae, with Solazyme feeding sugar to bugs in the dark and completely skipping photosynthesis. (Hmm...I wonder what sort of selection pressure that is putting on their algae strains?)  If Amyris does use algae -- sorry, when Amyris starts using algae -- the company will almost certainly be using it as a feedstock fed to microbes that then produce fuels.  This would require building a front-end process onto their yeast production system, but I don't see that as taking very long to happen.  See my earlier post on Blue Marble Energy.

Things are moving forward.  I would note that I see a lot of stainless steel in the photos of Aymris' pilot plant.  I am no fermentation jock, but it seems that they could probably use solvent resistent plastic as their culture vessels.  Here is one home-brew kit that basically just consists of plastic buckets.  Maybe that is a step for later.

Congratulations to everyone at Amyris.  Keep up the good work.

iGEM 2008: Surprise -- The Future is Here Already.

I'm back from a weekend at MIT serving as a judge for the International Genetically Engineered Machines Competition.  Here are a few thoughts on the competition.

The "international" flavor continues to strengthen.  Of the six finalists, three were from the U.S., two from Europe, and one from Asia.  There were 85 teams registered, almost all of whom showed up.  I was hoping for more biofuels/energy projects, but perhaps that fad is already past.

The top three teams were (here are the full results): 1) Slovenia 2) Freiburg 3) Caltech.

First, a couple of slightly blurry iPhotos (when the hell is Apple going to upgrade that camera?):

IMG_0138 Tom Knight receives the BioBrick from the 2007 winner, Peking University.

IMG_0140 A collective dance party while the competitors wait for the judges.

IMG_0141 Tom Knight awards the BioBrick to the 2008 winners, Slovenia.

Several of the 2008 projects implement ideas that have appeared in science fiction stories and in my own speculations about the future of biological technologies:

UCSF characterized a fusion protein that enables epigenetic control of gene expression through chromatin silencing.  This, in effect, gives the user (which could be the cell itself) a new control knob for building memory circuits in eukaryotes.  I seem to recall that this is the basic innovation in Greg Bear's Blood Music that brings about the end of the world through Green Goo.  Go UCSF!

Caltech and NYMU-Taipei (check out the killer Wiki) both modified commensal E. coli strains to serve as therapeutics.  Caltech built a bunch of new functionality into the probiotic strain Nissle 1917, including microbicidal circuits, Vitamin B supplements, and lactase production (big kudos to Christina Smolke, here).  Taipei built a "Bactokidney" for people with kidney failure: cells that attach to the lining of the small intestine and absorb nasty substances that would otherwise need to be removed via dialysis.  These are both very cool ideas.

Seeing these projects brought back shades of a scenario published in Bio-era's "Genome Synthesis and Design Futures: Implications for the U.S. Economy".  (I wrote the original story, which was less complicated but slightly more nefarious than the Bio-era version, in 2005 as a short, provocative piece of a larger report for a TLA -- a three letter agency.)  Almost all the technology described below has been published in bits and pieces -- fortunately, it has not yet been put together in one microbe.

In 2008, the North Korean government launches a secret program to develop biological tools that can be used to pacify target populations for crowd control or military purposes. North Korea's research draws on Soviet work on modifying pathogens to express mood-altering peptides, and the demonstration by U.S. scientists at the National Institutes of Health that common commensal strains of E. coli could be modified to secrete specialized peptides in human intestines.  Modifying the same strain used by the NIH, available in an over-the-counter probiotic pill, the North Koreans secretly produce an organism that produces peptide hormones easily absorbed through the intestinal wall.

With further modifications to allow the peptides to enter the brain, the new strain produces a calming, almost sedative, effect on colonized individuals. Combined with a genetic circuit that confers both antibiotic resistance and upregulation of the peptides upon exposure to a chemical that can be dispersed like teargas, these modifications enable the government to pacify crowds in times of crisis. The E. coli can be distributed via food and water to target populations.

To maintain the presence of the genetic circuit within the population, the new strain is equipped with an antibiotic resistance mechanism from V. cholera that causes plasmids containing the entire genetic circuit, including the regulatory genes and the mood modification genes, to be horizontally transferred to other bacteria upon treatment with common antibiotics.

In 2009, Pyongyang uses military forces to suppress a widening political uprising against the regime. Reports of a "pacifying gas" quickly emerge, raising allegations about the use of chemical weapons. U.S. intelligence agencies claim that North Korea has used a novel combination of biological and chemical weapons against rioters, leading the U.S. to declare that Pyongyang has violated the international treaty on bioweapons. Pacifist biohackers undertake to recreate the microbe , or to invent new versions to use as "peace weapons" against armies.

When a U.S.-led coalition attempts to impose an economic embargo against North Korea, the Chinese government uses its military to secure supply lines to North Korea. A military standoff between U.S. and Chinese forces ensues.

Here is the original inspiration: "Toward a live microbial microbicide for HIV: Commensal bacteria secreting an HIV fusion inhibitor peptide". (I'd completely forgotten that I blogged the original paper.)

Slovenia won (again) with "Immunobricks" by engineering new vaccines. The technology they used forms the basis of arguments about rapid, distributed vaccine production we made in Genome Synthesis and Design Futures (Section 4.3, in particular), which I've also written about extensively here on this blog, and which will show up in my book.  Yet all of a sudden it's real, all the more so because it was an iGEM project.

From Slovenia's Wiki abstract:

Using synthetic biology approaches we managed to assemble functional "immunobricks" into a designer vaccine with a goal to activate both innate and acquired immune response to H. pylori. We successfully developed two forms of such designer vaccines. One was based on modifying H. pylori component (flagellin) such that it can now be recognized by the immune system. The other relied upon linking H. pylori components to certain molecules of the innate immune response (so called Toll-like receptors) to activate and guide H. pylori proteins to relevant compartments within the immune cell causing optimal innate and acquired immune response. Both types of vaccines have been thoroughly characterized in vitro (in test tubes or cells) as well as in vivo (laboratory mice) exhibiting substantial antibody response. Our strategy of both vaccines' design is not limited to H. pylori and can be applied to other pathogens. Additionally, our vaccines can be delivered using simple and inexpensive vaccination routes, which could be suitable also in third world countries.

If you've read this far into the post, you should definitely spend some time on Slovenia's Wiki.

Here's the short, pithy version: There is presently no vaccine for H. pylori.  Between June and October this year, seven undergraduates built and tested three kinds of brand new vaccines against H. pylori.  (They also put a whole mess of Biobrick parts into the Registry, which means those parts are all in the public domain.)

Yes, yes -- it's true, getting something to work in a mouse and in mammalian cell culture is a long way from getting it to work in humans, or even in ferrets.  But the skill level and speed of this work should make everyone sit up and take notice.

So it is worth pondering the broader implications of these projects.

The Slovenian team clearly has access to very high quality labs and protocols.  Mammalian cell culture can be very fiddly unless you know what you are doing and have the right equipment (I speak from painful experience, lo those many years ago in grad school).  The Caltech and Taipei teams also clearly have a great deal of support and mentoring.  Yet while bashing DNA and growing E. coli are not particularly hard, the design and testing of the coli projects is very impressive.

Despite all the support and money evident in the projects, there is absolutely no reason this work could not be done in a garage.  And all of the parts for these projects are now available from the Registry.

Over the past couple of years, in various venues, I have tried to point out both the utility and inevitability of proliferating biological technologies.  iGEM 2008 drives home the point yet again.  In particular, the ability to rapidly create vaccines and biological therapeutics points the way to increased participation by "amateurs", whether the professionals (and policy makers...and security types) are ready or not.  I'm also thinking back to "peer reviews" in which I was excoriated for suggesting this kind of work was within the reach of people with minimal formal training.  Because, really, you need a PhD, and an NIH grant, and tenure, to even think of taking on anything like a synthetic vaccine.  Oh, wait...

Although I've predicted in writing that this sort of thing would happen, I frankly expected practical implementation of both the rapid, synthetic vaccines and the modified commensal bacteria to take a few more years. Yet undergraduates are already building these things as summer projects.

It didn't really hit me until I started writing this post earlier this afternoon, but as I ponder the results from this year's iGEM only one thought comes to mind: "Holy crap -- hold on to your knickers."

The world is changing very, very quickly.

Heading to MIT and iGEM

I'm off to give a talk at MIT on Friday and be a judge for iGEM again this weekend.

Here are talk coordinates (I've no idea how big the room is):

"Engineering (and) the Bio-economy"
Rob Carlson

7 November, 2008
12 to 2 PM
MIT
E38, 6th floor conference room.
292 main street, directly above the MIT press book store.

Can't wait to see what the students have come up with this year.

Update on Influenza Subtype Activity in 2007-2008

While addressing some comments from Ralph Baric on one chapter my book, I had reason to go find statistics on influenza subtype activity last year.  Those interested in keeping up on recent flu activity should peruse this July, 2008, report from the CDC: Influenza Activity --- United States and Worldwide, 2007--08 Season.

Here is the breakdown on subtype activity:

During September 30, 2007--May 17, 2008, World Health Organization and National Respiratory and Enteric Virus Surveillance System collaborating laboratories in the United States tested 225,329 specimens for influenza viruses; 39,827 (18%) were positive. Of the positive specimens, 28,263 (71%) were influenza A viruses, and 11,564 (29%) were influenza B viruses. Among the influenza A viruses, 8,290 (29%) were subtyped; 2,175 (26%) were influenza A (H1N1), and 6,115 (74%) were influenza A (H3N2) viruses. The proportion of specimens testing positive for influenza first exceeded 10% during the week ending January 12, 2008 (week 2), peaked at 32% during the week ending February 9, 2008 (week 6), and declined to <10% during the week ending April 19, 2008 (week 16). The proportion positive was above 10% for 14 consecutive weeks. The peak percentage of specimens testing positive for influenza during the previous three seasons ranged from 22% to 34% and the peak occurred during mid-February to early March. During the previous three influenza seasons, the number of consecutive weeks during which more than 10% of specimens tested positive for influenza ranged from 13 to 17 weeks.

Of note, 26% of samples positive for influenza were the H1N1 subtype -- the same as the 1918 flu -- which means we all have probably been exposed to it and have some immunity.  That does not mean the particular combination of genes in the 1918 flu would be harmless if it showed up again, but rather than our immune systems should be able to better recognize that bug and thus might defend against it better than the first time around.

Land Reform in China

The IHT is carrying news of a new land policy in China.  Here is the lead:

Chinese leaders approved on Sunday a policy that will in theory allow peasants to buy and sell their land rights, a move that sets in motion the nation's biggest economic reform in many years, according to a report by Xinhua, the state news agency.

The print version I picked up in Hong Kong today is a bit different.  It carries this crucial bit of information:

The government's goal is to double the per-capita disposable income of rural residents by 2020 from the 2008 level, according to Xinhua.

It is unclear how much the plan is intended to increase total per capita income in rural areas.  I think this is particularly important because it will strongly influence how much almost 800 milllion people (according to the article) have to spend on food.  Implementing the land reform plan may put a time scale on the increase in food demand that I speculated about recently in "More on China's Economy, Food Production, and Food Demand".

If the numbers in that post are mostly correct, this would mean that if China is going to stay self-sufficient with respect to food supply needs to increase its domestic production by something like 20% in the next 11 years.  They have their work cut out.

Synthetic Biology 4.0 – Not so live blog, part 1

What a difference a few years makes.  SB 1.0 was mostly a bunch of professors and grad students in a relatively small, stuffy lecture hall at MIT.  SB 2.0 in Berkeley expanded a bit to include a few lawyers, sociologists, and venture capitalists.  (I skipped 3.0 in Zurich.)

At just over 600 attendees, SB 4.0 is more than twice as big as even 3.0, with just under half the roster from Asia.  The venue, at the Hong Kong University of Science and Technology, is absurdly nice, with a view over the ocean that beats even UCSB and UCSD.  Kudos also to the organizers here.  They worked very hard to make sure the meeting came off well, and it is clear they are interested in synthetic biology, and biotech in general, as a long term proposition.  The Finance Minister of Hong Kong, John Tsang, spoke one evening, and he was very clear that HK is planning to put quite a lot of money and effort into biology.

Which brings me to a general observation that Hong Kong really cares about the future, and is investing to bring it along that much sooner.  I arrived a day early in order to acclimate a bit and wander around the city, as my previous visit was somewhat hectic.  Even amid the financial crisis, the city feels more optimistic and energetic than most American cities I visit.

I will have to write up the rest of the meeting when I get back to the States later this week.  But here are a few thoughts:

As of the last few days, I have now seen all the pieces necessary to build a desktop gene printer.  I don’t have prediction when such a thing will arrive on the market, but there is no doubt in my mind that it is technically feasible.  With appropriate resources, I think it would take about 8 weeks to build a prototype.  It is that close.

Ralph Baric continues to do work on SARS that completely scares the shit out of me.  And I am really glad it is getting done, and also that he is the one doing it.  His work clearly demonstrates how real the threat from natural pathogens is, and how poorly prepared we are to deal with it.

Jian Xu, who is better known for his efforts to understand the human gut microbiome, spoke on the soup-to-nuts plant engineering and biofuels effort at the Qingdao Institute of Bioenergy and Bioprocess Technology, run by the Chinese Academy of Sciences (QIBEBT).   The Chinese are serious about putting GM plants into the field and deriving massive amounts of energy from biomass.

Daphne Prauss from Chromatin gave a great talk about artificial chromosomes in plants and how they speed up genetic modification.  I’ll have to understand this a bit better before I write about it.

Zach Serber from Amyris spoke about on their biofuels efforts, and Amyris is on schedule to get aviation fuel, diesel, and biogasoline into the market within the next couple of years.  All three fuels have equivalent or better characteristic as petro-fuels when it comes to vapor pressure, cloud point, cetane number, octane, energy density, etc.

More soon.

Off to Synthetic Biology 4.0 in Hong Kong

Tomorrow I am off to SB 4.0 in Hong Kong.  I will be leading a lunchtime workshop on "Commercialization of Genetic Parts".

Here is the abstract:

Synthetic Biology presently finds most of its participants in academia. Yet many applications, for example fuels, drugs, and materials, will be developed and used within a very different context than universities and other non-profit organizations. How can we spin up an economy based on composable parts that people can actually use to pay the rent in one way or another?

I expect the direction of the discussion will be guided mostly by who actually shows up.  If we wind up talking about Biobrick Parts (TM), then we will need to hear more about whatever license the Foundation is trying to put together.  My understanding is that the details of this license are still under wraps, so the discussion could well focus instead on how one can commercialize any widget these days.  If that is the way it goes, I'll be happy to share my experiences in exploring the possibilities of garage biology.  Here is the short version: Building a prototype is not so bad, and contract manufacturing is a royal pain in the ass to get going.  So it's just like the rest of the economy...

Assuming we have decent web access, I will be blogging during the meeting.  Stay tuned.

Company Profile: Blue Marble Energy

A couple of months ago I met the founders of Blue Marble Energy at a party for the Apollo Alliance.  Following up, I sat down with the CEO, Kelly Ogilvie, to learn about Blue Marble, which is the only "algal biofuel" company I have come across that really makes sense to me.  (While at the party, I also chatted with Congressman Jay Inslee for quite a while.  Smart fellow.  Anyone interested in energy policy should have a look at his book, Apollo's Fire: Igniting America's Clean Energy Economy.)

Full disclosure: Blue Marble and Biodesic may begin collaborating soon, so I am not an entirely disinterested observer.

Blue Marble Energy is built around the idea of "recombining" existing biological processes to turn biomass into valuable products.  From the website: "[Blue Marble Energy] uses anaerobic digestion to generate natural gas and other valuable bio-chemical streams."  The company is distinguished from its competitors by its focus on using micro- and macro-algae harvested from natural blooms, including those caused or enhanced by human activity, as feedstock for artificial digestion systems modeled on those of ruminants.  Blue Marble combines different sets of microbes in a series of bioreactors to produce particular products. 

In other words, Blue Marble is using industrialized, artificial cow stomachs to produce fuel and industrial products.

The company's general strategy is to first digest cellulose into synthesis gas (carbon dioxide and hydrogen) using one set of organisms, and then feed the synthesis gas to organisms that generate methane or higher margin chemicals and solvents.  The company expects to produce 200-300 cubic meters of methane per wet ton of algal feedstock.  While biofuels are an obvious target for technology like this, the company also recognizes that fuels are a low margin commodity business.  Thus Blue Marble also plans to produce higher margin industrial products, including solvents such as various esters that sell for $400-800 per gallon.

While other companies are attempting to directly produce fuels from cultured algae, Blue Marble believes these efforts will be hampered by growth limitations in most circumstances.  Biofuel production from algal lipids synthesized during photosynthetic growth requires conditions that cause metabolic stress, resulting in lipid production, but that also limit total biomass yield to ~2-5 grams per liter.  In contrast, Blue Marble "respects the complex ecology", in the words of Mr. Ogilvie, and relies on photoheterotrophic growth of whatever happens to grow in open water.

Blue Marble has already obtained contracts to clean up algal growth caused by human activity around Puget Sound.  The company typically harvests ~100 grams per liter from these "natural" algal blooms.  Future plans include expanding these clean up operations around the U.S. and overseas, and growing algae in wastewater, which would provide a high-energy resource base for both closed and open system growth.  In principle, because the technology is modeled on ruminant digestion, many different sources of biomass should be usable as feedstock.  Experience thus far indicates that feedstocks with higher cellulose content result in higher yield production of fuels and solvents.

Compared with other algal biofuel companies, Blue Marble does not presently require high capital physical infrastructure for growing algae.  However, the company will rely on marine harvesting operations, which bring along a different set of complexities and costs.  I wonder if the company might be best served if it outsourced harvesting activities and focused on the core technology of turning biomass into higher value products.

While the Blue Marble is not now genetically modifying their production organisms, this will likely prove a beneficial move in the long term.  Tailoring both the production ecosystem and the metabolisms of component organisms will certainly be a goal of competitors, as is already the case with companies spanning a wide range of developmental stages, including DuPont, Amyris, and Synthetic Genomics.  Yet whereas modified production organisms grown in closed vats are likely to face little opposition on any front, genetically modified feedstocks grown in open waters are another matter.  For the time being, Blue Marble has an advantage over plant genomics companies because in the company's plans to use unmodified biomass as feedstock, whether algae or grasses, it will avoid many regulatory and market risks facing companies that hope to grow genetically modified feedstocks in large volumes. 

They have a long way to go, but in my judgement Blue Marble appears to have a better grasp than most on the economic and technical challenges of using algae as feedstock for fuels and materials.

Further reading:

"It came from the West Seattle swamp - to fill your tank", Eric Engleman, Puget Sound Business Journal, August 8, 2008

"Swamp fever", Peter Huck, The Guardian, January 9 2008
http://www.guardian.co.uk/environment/2008/jan/09/biofuels.alternativeenergy

"New wave in energy: Turning algae into oil", Erica Gies, International Herald Tribune, June 29, 2008

Methane Time Bomb Update

Following on its coverage of an expedition to Russia's northern coast that found methane deposits leaking through melting permafrost into the water and atmosphere, The Independent has news that a British expedition to the seas off the coast of Norway has discovered "hundreds of methane plumes".  From the article:

Yesterday, researchers on board the British research ship the James Clark Rosssaid they had counted about 250 methane plumes bubbling from the seabed in an area of about 30 square miles in water less than 400 metres (1,300 feet) deep off the west coast of Svalbard. They have also discovered a set of deeper plumes at depths of about 1,200 metres at a second site near by.

The story notes that "It is likely that methane emissions off Svalbard have been continuous for about 15,000 years – since the last ice age."  I think it is fascinating that these plumes have only just been discovered.  This means the methane budget of the atmosphere is probably still quite poorly understood, even as it is clear new sources of methane are opening due to climate change.