The Bowman Lab

When I started graduate school I new absolutely nothing about computing on anything more high performance than a laptop.  I assumed that clusters and servers were exclusive to large, well-funded labs.  Access to these items is a huge limiting factor in almost any field of research.  This is certainly the case in microbiology where it is increasingly difficult to close the loop on an interesting story without volumes of sequence data.  At the time (and for a long time after) I was indignant at this seeming injustice – only large labs with large budgets had access to the large computers required for the most compelling investigations (the results of which enable additional funding…).  Over time I’ve slowly come to appreciate that this is not the case.  A high performance computer is cheaper than many items deemed essential to a microbiology lab (-80 C freezer (~ $10 K), microscope (~ $20 K)).  I suspect there are other students out there with similar misconceptions, so here is a short compilation of things I’ve learned regarding research computers.  Imagine that you are a new faculty member, or a postdoc with a little fellowship money, looking to purchase a computer for basic bioinformatics (e.g. genome assembly and annotation).  What will you have to spend?

A necessary consideration is how you choose to use a computer in a research setting.  Our lab’s “server” is a MacPro workstation (more on that shortly).  After some trial and error I settled into an arrangement where I work almost exclusively from my laptop.  I use ssh and scp to access the server, and almost never sit in front of it.  The rest of this post assumes that the purchaser works in a similar fashion.  It’s a matter of personal style, other lab members like to park in front of the monitor (I’d rather park in the coffee shop, or at least my office).  Unlike almost everyone else in the world of bioinformatics my laptop is a PC.  Again, a matter of style.  It works great for me.  The point is that it works almost seamlessly with the MacPro, and all other Unix/Linux servers.  It really doesn’t matter what you connect to it with, the server is the important thing.  It’s worth bringing this up because, as a graduate student, I’d rather have a cheap laptop and a decent remote workstation than a nice laptop (of course in an ideal world you would have both).  If you can’t ask your advisor for both, ask for the workstation…

Our MacPro was purchased back in 2008, for all the right reasons.  Next generation sequencing, which generates orders of magnitude more data than older methods, had just been introduced.  A Mac with 8G of RAM and 8 cores seemed like more than enough computer.  Nobody in either of the labs using the computer knew much about Unix or Linux, so it didn’t seem like there was much of a choice.  Unfortunately in this decision a cardinal rule of computers was violated – a computer deemed sufficient to do the job you are purchasing it for is likely to be insufficient for the next job.  It’s hard to justify spending  money on additional resources without a clear use for them, but it just has to be done in some cases.  When I started working with next generation data I upped the RAM to 24 G and added an additional 2T HD.  Theses upgrades make the MacPro suitable for 75 % of the tasks I need to perform on a day to day basis, but the other 25 % is kind of essential…

The top of the line MacPro today, with 8T of storage, 12 cores, and 64 G of RAM will run you around $6,600 (tower only).  For a long time I assumed this was the only way to go – more capable workstations must be prohibitively more expensive, right?  Since there’s a lot that you can’t do with only 64 G RAM and 12 cores I assumed that that work was the sole purview of well funded labs.  Sometime back I explained this dilemma to a colleague in our geology department who deals with large geospatial datasets, and who clued me in on the price of computers.

Mac’s are friendly,  but some friends can be bossy and expensive to hang out with.  In this case it takes a bit of tinkering to make a Mac work well as a Unix server.  If that’s how you’re using it than you aren’t using the features that you’ve paid a lot of money for – namely the Mac OS GUI front end.  True Unix/Linux workstations are comparatively cheap.  For ~ 5 K I found two local companies who could provide a server-grade tower with 128 G RAM (minimal for many applications, but expandable to 512 G), and 2 x 8 Intel Xeon processors (16 cores total).  There’s plenty that such a machine can’t do, but it’s not a bad start.

I know that other groups have migrated to Amazon ECS for research computing, but there are clear downsides.  Amazon “instances” are expensive, and limited to 64 G RAM.  If you aren’t a computer expert you could spend a lot of money just trying to figure out how to make things work.  It also seems that data exploration would be limited by cost, and by the slight difficulty inherent in setting up instances.  Most research institutions have some kind of high performance computer environment; I’ve used the University of Washington’s Hyak cluster for a lot of my work.  This environment is great when many processors are desired, but, as with Amazon ECS, it is expensive and memory limited.  I’m pretty new at this and just getting a sense of what’s out there.  If you’ve had a good or bad experience with any of this technology leave a comment!

As an offshoot from one of my dissertation projects I’m developing a simple model to explore protein evolution.  The model takes an amino acid sequence and mutates it, selecting for mutations that result in improvement for one of several measurable protein parameters (flexibility, hydropathy, aliphatic index, isoelectric point, or aromaticity).  I’m using it to explore how selection for one parameter results in unintended changes to the other parameters.  I implemented the model in Python but I’m using R for data analysis and plotting.  For each run I have several hundred measures of the five different protein parameters.  There are a few ways to visualize this, but since the dataset represents a temporal progression it makes sense to create an animation of the output.  After some exploration this is the method that I came up with.

Assuming that you have some table of data (evol_param_d_matrix_1 in this example) in R for which a row represents a single time point, you can generate a figure (I chose a barplot) for each time point as such:

jpeg(filename = 'animations/Rplot%03d.jpg')
for (i in 1:length(evol_param_d_matrix_1[,1])){
  barplot(evol_param_d_matrix_1[i,],
          ylim = c(-0.5, 1.5),
          axisnames = T
          )
}
dev.off()

Note that it is extremely important to specify ylim (and xlim, if this axis might vary for your data) to prevent the scale of the plot changing from image to image.  The above code saves each plot (one per line in evol_param_d_matrix_1) as a jpeg to the directory “animations”, labeled Rplot000.jpg, Rplot001.jpg, etc.  To convert this sequence of still images to video I used the free tool (for linux/mac) ffmpeg.  For mac install using homebrew.  I called ffmpeg using:

ffmpeg -r 15 -qscale 1 -i 'Rplot%03d.jpg' test.mp4

That command tells ffmpeg to use 15 frames per second, with a quality of “1” (the highest quality), on all jpegs matching the given naming pattern.  I first tried to use Rplot*.jpg, but this ordered the frames alphanumerically, rather than numerically. The output file is of course test.mp4.  Changing to test.avi would instead output in that format.  For reasons unknown to me, neither the command tee nor the redirect could capture ffmpegs screen output to a logfile, which might have been useful for the initial troubleshooting.

The ffmpeg documentation is, unfortunately, rather dense and written for people with a lot more audio/visual background than I have.  It looks like ffmpeg is a commonly used tool however, so there are quite a few easy(ish) to follow tutorials online.  Here’s the final product:

We just finished a crazy few days of sampling and I’m only now, back in Seattle, getting a moment to assimilate all of it.  The Icy World team from JPL arrived late in the day on May 1, and we wasted no time getting started with field work.  The primary mission was to re-sample a number of shallow, methane rich lakes that the team had sampled in October, just as the lakes were starting to freeze.  What we quickly discovered at the first lake we visited however, was that a particularly cold winter and unusually thin snow cover had frozen most of the lakes straight through to the bottom.  This made it impossible to recover instruments deployed in the fall or to sample lake water.  Fortunately not all of the team’s work depended on lake water.  We recovered ice cores from several lakes and measured the flux of methane from the boreholes.

No water in this lake.  Kevin Hand and Andrew Klesh prepare to section an ice core that extends all the way to the lake sediment.

The ice from these lakes is very interesting.  As it forms it traps bubbles of methane; the ice looks like a frozen glass of champagne.  If you find a spot directly over an active methane “seep” the bubbles can be huge.  In one spot our drill broke through to a methane pocket large enough to choke out the gas powered drill motor.  We gave that hole the “lighter test” – a borehole with a high methane concentration will actually catch fire.  This one burned intensely for around five seconds!

Photo courtesy of the JPL Icy Worlds project. This photo was taken while we were working at the ice edge. I like it because it shows something I hadn’t considered before – ice algae growing on the top of rafted sea ice. Normally we think of growth as occurring only on the bottom of sea ice. There is a lot of rafted ice along pressure ridges however, perhaps we’ve been underestimating the space available for ice algae to grow.

After two days of sampling lake ice we switched things up and returned to the ice edge near where I had sampled frost flowers on April 30.  The lead was open again and the whaling crew had their boats in position for the first whales of the season.  It was extremely kind of them to allow us temporary use of their forward whaling camp – the Bowhead whales started to appear while we were there and they could have viewed our presence as a distraction or intrusion (for the whales and the whalers).  In addition to collecting more frost flowers and taking some spectacular images of the ice edge it was great to check out the whaling operation up close.  The hunters really have whale behavior dialed.  When the lead opened we were given a prediction of when the whales would arrive – which was accurate to within a couple of hours, and were told that the Beluga whales would lead the larger Bowheads by about two hours.  Not long after the Belugas started swimming past us at a leisurely pace we saw Bowheads breaching in the distance!

Lewis Brower, ready for the hunt to begin. Lewis Brower first introduced me to sea ice in 2009. Many thanks Lewis, and good luck this season! I was told that the hide boat that he is in is around 60 years old – that boat has seen a lot of whales and a lot of change to Barrow.

It’s coming right for us! A Beluga dives under the ice edge. Not long after the Belugas started passing us we could see Bowhead whales breaching in the distance.

The next day (now starting to show a little exhaustion) we returned to the lakes, this time finding some with a little water left in them.  In addition to ice and water samples the team collected some amazing footage of the under-ice lake ecosystem, including a couple of sizable whitefish shown in the video below (Video courtesy of the Nasa Astrobiology Institute’s Icy Worlds project at JPL).  When I see a lot of biomass like that (also represented by the sizable microbial mats in the lake) I think energy.  The interesting thing that this video illustrates to me is the amount of energy that can be contained in a shallow water column underneath a lot of ice – though of course these lakes are most energy rich in the summer, when they are well oxygenated and full of phytoplankton fixing carbon.  In the winter however, the flux of methane to these lake should serve as a continued carbon source – with methanotrophic bacteria feeding small protists, which are fed on by larger lake organisms.  That’s a testable hypothesis and one that would be worth following up on.

One more day of lake sampling (and a couple of long nights in the lab) saw us to the end of the field effort.  It was one of my most memorable trips to Barrow, and a very successful sampling effort.  Many thanks to our guides at Umiaq, and to Icy Worlds for letting me tag along…

I’ll be taking off for Barrow in just a few hours, the last field effort of my dissertation – and my last chance at collecting frost flowers!  This will be my 6th trip up in the last four years, although two of those trips were pretty short, when conditions made work impossible or impractical.  Here’s a short photo summary of our previous Barrow efforts.

April, 2009. My first time working on sea ice, and our first frost flower samples that made it into a publication (Bowman and Deming, GRL, 2010).  Many thanks to Hans-Werner Jacobi and Florent Domine for the invitation to participate in their field effort.

Our first serious, NSF funded effort to look at the microbiology of frost flowers, in February, 2010. This was a difficult trip, dominated by pretty extreme weather. There was no open water anywhere, and whenever we tried to open an artificial plot to grow frost flowers it would quickly be covered by blowing snow.  A major blizzard pinned us down towards the end of the trip, the airport was closed for days and the road from the research station to town became impassible.  At one point, with our truck stuck in a snowdrift, my lab mate Jesse and I watched snow drifts build around the truck faster than we could dig them down (we eventually abandoned the truck, fortunately we weren’t too far from the station).

Following our failed effort in February of 2010, we received an email in April that natural frost flowers were forming. I went up the next day and sampled from this lead, these are the samples in our 2013 paper (Bowman et al., EMIR, 2013). That night it snowed a foot, eradicating all frost flowers. If I’d been a day later I would have missed them.

2011, another tough year. Ice conditions that winter were extreme, and the whaling crews were having a very difficult time cutting trails through the landfast ice. I went up in March, which turned out to be much too early, and spent several frustrating days searching for a path out to the lead. We covered a lot of ground and found nothing other than these (very blurry) polar bears!

After a few days of searching I went home, and came back in April after the ice conditions had improved. This massive wall of ice shows how bad the weather was that year. It’s a pressure ridge created by the wind stacking pans of ice on top of one another, then sheared in half by a major storm. We did end up finding frost flowers, sort of. Everything was just weird that year. No new frost flowers were growing where we could reach them (the wind keep blowing them to places we couldn’t go), and the ones we could reach were very low in salinity.

So there we are… three field seasons over four years, five trips, and two publications (we are working on a third from the 2010 samples, the metagenomics project that I’ve written about here).  I’ve been nervously watching the weather for the last few days and am cautiously optimistic.  Right now it’s pretty warm (18F/-8C) and snowing lightly, not good, but tomorrow it should be cold and precipitation free.  I’d really like to find some week-old frost flowers, which isn’t going to happen with today’s snow, but the chances of finding new ones are good.

I’m really interested in connecting the unexpected microbial community that we observed in 2010 to the wider environment however, which means that even if I don’t find frost flowers this field season can be a success.  I’ll be sampling open water, young sea ice, and even lake water (working with colleagues on a NASA project who will arrive a few days after me) – environments that are much less dependent on conditions than frost flowers!

But you can grow flowers there. As readers of this blog know, one of the Deming Lab’s major research directions is the microbiology of the sea ice surface – frost flowers, saline snow, and related features.  Since the sea ice surface changes over time (collecting fresh snow, among other things) the best time for a microbiological investigation is when the ice is relatively young.  Of course this presents many challenges.  In addition to being thin – and therefore difficult to travel over – the formation of young sea ice is difficult to predict.  An Arctic bay that hosted a beautiful field of frost flower during a given month one year might be open water, or 50 cm thick ice, the next!

In 2012 Jody Deming traveled to Daneborg in northeast Greenland to take part in an exploratory field effort with colleagues in the Arctic Science Partnership (ASP).  Among other things they were trying to assess the potential for Daneborg as a long-term study site, and trying to develop methods for bringing together the tools and techniques of many different disciplines to better understand the sea ice ecosystem.  There is a polynya (a place where the ice remains thin, or even non-existent, throughout the winter) in the fjord at Daneborg, an ideal situation for studying young sea ice.  The team was thwarted however, by weather and ice conditions.  Unable to reach the natural polynya they made their own (check out this post to see what happened when we tried this in Antarctica)!

This video from Aarhus University in Denmark has some great footage of the artificial “pond”, and a short interview with Jody, talking about some of her research interests.

Frostflower from Aarhus Universitet on Vimeo.

I got an email the other day from a colleague who teaches science at Soap Lake High School, a rural high school in eastern Washington that, in my opinion, punches above its weight in the sciences.  One of his students is preparing a presentation on Europa, and she had some questions for a Europa expert.  I’m not a Europa expert, but I interact with people who are, so I suppose I’m only one degree removed from actual knowledge of Europa…

Soap Lake High School is located (predictably) near Soap Lake; a highly alkaline, highly stratified saline lake.  Because of the lake’s physical structure and chemistry it is a useful analogue for the Europan ocean.  It’s not Lake Vostok, but no analogy is perfect.  Europa is a pretty hot topic in astrobiology, and there is plenty of information out there on it, but I figured it wouldn’t hurt to post my answers to her questions here.  It also lets me expand on the original answers a little.

Why do scientists think that life could be on Europa?

Scientists think that life needs two things – liquid water and a source of energy.  Europa is one of the few places in our solar system (other than Earth) where both of these things are likely abundant.  There is a lot of evidence that underneath Europa’s ice shell is a huge ocean, much larger in volume than the oceans on Earth.  Because Europa is so close to Jupiter, and because of Jupiter’s massive size, there is a huge gravitational force exerted on Europa.  This gravitational force heats Europa much in the way that Earth is heated by radioactive decay in its core, possibly enough to produce low-level volcanism at the sea floor.  As on Earth, seafloor volcanism  produces springs, black smokers, and other sources of reduced elements, useful for life as an energy source (when paired with oxidized material).  On Earth these volcanic sites are hotspots for life.

For energy, life requires both oxidized and reduced material.  Oxidized material should be abundant at the surface of Europa’s ice shell, thanks to constant bombardment by radiation.  There is lots of evidence that Europa’s ice shell turns over gradually, bringing oxidized material into the ocean and exposing fresh ice to radiation.  On Earth sea ice is a hotspot for photosynthetic life because sunlight is abundant – like blacks smokers it is energy rich.  A similar situation could exist on Europa at the interface between ice and water, since this environment is optimally located between the oxidized ice and the reduced ocean.  If Europa were a battery the seafloor and ice surface would be the negative and positive terminals (respectively).  Life at the interface acts as a circuit, using the electrons for biological processes as they flow from the negative to the positive terminal (from seafloor derived material to ice surface derived material).

From Seeking Europa’s Ocean (Pappalarado, 2010). This cut away shows many putative features of the Europan ice shell, including the strong temperature gradient, convective diapirs, and the surface features produced by these internal processes.

How is Soap Lake similar to Europa?  How is it different?

There is a lot of “chemoautotrophic” life in Soap Lake.  This is life that, rather than using the sun as a primary energy source (in the case of plants by creating reduced carbon and free oxygen), uses reduced and oxidized chemicals – the negative and positive battery terminals discussed before.  These chemicals may have a biological source (for example ammonia or nitrate or organic carbon) or a geological source (like iron).  If the chemical being used is not organic (does not contain carbon-hydrogen bonds), life is said to live lithoautotrophically – this kind of life can theoretically happen in the absolute absence of sunlight (no photosynthesis going on anywhere in the environment).  Finding locations on Earth where this is happening is challenging, because photosynthesis influences almost everything here.  On Europa there is essentially no light that can be used in photosynthesis, so we would expect a lot of lithoautotrophy.  Without photosynthesis the ecosystem of Soap Lake would look pretty different, but there are good examples of microbial life there that can serve as a model for life on Europa.  One example is bacteria that gain energy via sulfur oxidation, where reduced sulfur compounds are oxidized by oxygen.

If there is life of Europa, what do you think it looks like?

I think it would look pretty much like life on Earth – or at least microbial life from the right habitats on Earth.  Evolution seems to produce the same solutions to problems over and over again – this idea is called convergent evolution – and the problems life would face on Europa are not so different than what it faces on Earth.  It seems likely that evolution would produce the same basic structures and metabolic strategies to deal with these problems.

Do YOU think there’s life on Europa, and what will it mean if there is?

I think Europa is the most likely place in the solar system, other than Earth, for life to exist.  The fact that it is so far from Earth (unlike Mars) makes this even more exciting.  If we find life on Europa there is almost no chance that it came from Earth, it would have to have had a separate origin.  This would tell us a lot about how widespread life is in the universe.  Finding life on Mars would be a little different, chances are that it would have something to do with life on Earth (Earth and Mars exchange asteroids and other material all the time, providing a vector for biological contamination).

How soon do you think it will be before we know for sure if there’s life on Europa?

Unfortunately it will be a long time before we know if there is life on Europa – a couple of generations if we maintain an interest in exploring the outer solar system.  The most we’ve ever been able to study Europa came with the Galileo spacecraft which was sent to study the entire Jupiter system.  It flew past Europa multiple times and from those observations we know a little about the composition of the moon and the likelihood of liquid water.  We’ve never had a spacecraft in orbit around Europa however, let alone landed on its surface!  Since life is most likely to exist below several miles of hard, thick ice even getting to the surface won’t tell us if life exists.  Somehow we need to land on the surface and melt or drill through all that ice into the ocean below (warm ice rising in the shell might also contain life, but that would still involve a lot of drilling).  This is technically feasible but would be very very expensive to do.  We might find tantalizing evidence of life from a spacecraft in orbit or on the surface, perhaps in the form of chemicals that were most likely produced by life, but this isn’t absolute proof.