UT Austin's Lonestar supercomputer helps biologists save reef-building corals
Corals can genetically adapt to warmer waters from climate change, scientists say in a study that relied on bioinformatic analysis with supercomputers.
"Corals already have a genetic capacity to rapidly adapt to global warming," said Mikhail Matz, an associate professor in the Department of Integrative Biology at the College of Natural Sciences, The University of Texas at Austin.
Matz co-authored the coral study published June 2015 in the journal Science. "Basically, they already have the mutations which will save them from impending heat. It's only a matter of redistributing those mutations throughout the coral populations," Matz said.
Coral reefs around the world face a race against time. Globally, average temperatures of the surface sea waters where most corals live are rising.
The symbiotic algae called zooxanthellae that give coral their dark color convert sunlight to food. When waters become as little as one-two degrees Celsius too hot, zooxanthellae abandon coral and leave behind a 'bleached'-looking and severely weakened reef.
A prior 2014 study with computer models found that coral reefs have slowly evolved to keep one step ahead of deadly heat if their waters warm gradually. "The question is, do they have the capacity to adapt to this rapidly enough to keep up with the rate of global warming?" Matz asked.
Reefs can recover from short-term heat waves, but longer bouts can lead to widespread bleaching, such as the one exacerbated by an El Nino in 1998 that left 16 percent of the world's coral dead in its wake.
"Coral bleaching has become an increasingly frequent cause of coral death, with severe bleaching going on around the world right now," says Mark Eakin, coordinator of the Coral Reef Watch program at the National Oceanic and Atmospheric Administration.
Misha Matz and his collaborators began their study at the Great Barrier Reef in Australia. The reef has seen its share of troubles, losing an estimated 50 percent of its live coral since 1985, but it remains the largest coral reef in the world.
The Great Barrier Reef is so big it spans fifteen degrees of latitude, roughly covering the same area as Italy or Japan. And it dwells in waters that can vary four degrees Celsius (seven degree Fahrenheit) and even higher. Matz and colleagues sampled corals from one of the warmest locations on the Great Barrier Reed and cross-fertilized them with corals of the same species from a much cooler location, to see if they would pass their heat tolerance to their offspring.
It turned out to be a "logistical nightmare," Matz said. Coral only spawn once a year, so he and colleagues had to get it right the first time or wait another year to try again. "We were pretty lucky to get gametes from corals from different locations in our little buckets in the same time so we could cross them in different directions, different combinations, using sperm and eggs from one coral to cross-fertilize others from different populations, " Matz said.
From this one experiment Matz measured how larvae inherit their heat tolerance, using all-against-all crosses of coral taken from four colonies of the Great Barrier Reef at two locations, the warmer Princess Charlotte Bay and the cooler Orpheus Island.
"What we have found is that the heat tolerance of the coral babies that come out of these crosses can vary up to ten-fold, depending on whether their parents are coming from warmer or cooler locations," Matz said. This good news, Matz added, is the main take-home message from the study.
"Coral bleaching has become an increasingly frequent cause of coral death, with severe bleaching going on around the world right now."To dive deeper into the question of how corals adapt to heat, Matz measured which genes are working differently, or expressed, in coral larvae to resist heat. "Also, we looked for specific regions in the genome which are conferring resistance to heat by taking a thousand of coral larvae and basically cooking them up until almost all of them die and sampling the very last survivors — the hardest of the hard larvae, who are tolerating the heat best," Matt said.
His group compared the genetic composition of the 'hardest of the hard' across the genome to the rest of the larvae and found the regions in the genome that strongly correlated with heat tolerance.
"To boil this down to an actual answer of what biological processes are responsible for your feature of interest, this requires an intricate and multi-stage bioinformatic pipeline, much of it taken up by TACC computers," Matz said.
Matz called himself an 'old timer' when it came to utilizing TACC systems. Since 2009, his group used hundreds of thousands of CPU hours on the TACC Lonestar 2 and now-retired Ranger supercomputers to be first to sequence and analyze the transcriptome, or complete RNA, of Acropora millepora coral.
This latest coral study used about 386,000 CPU hours on Lonestar 4. "I'm working on Lonestar simply because I know it really well and the queues are shorter," he joked. "My tasks are not terribly computationally intensive. I don't need fast processors. I need a lot of computers working in the same time, even if they're not as super fast as they could possibly be," Matz said.
If you're a researcher interested in applying for allocations at TACC, check out this overview page.
"The computational resources of the Texas Advanced Computing Center played a huge role in actually enabling these analysis.""Our gene expression analysis used 54 samples, each one of them was close to one GB of data," Matz said. "Our genomic mapping pipeline used about 370 samples, approximately the same size. Just to process all that you do need big, nice fat computers."
Matz described himself as a self-starter, like many scientists who use supercomputers. "We managed ourselves so far," he said. An occasional trivial but show-stopping error such as accidentally deleted user profile did prompt a call though. "TACC people are extremely responsive and many of them reply in the middle of the night. Apparently geeks never sleep. That's very good," Matz said.
Part of the work that Matz's lab did was to show others how to fish for gene data pearls themselves. They developed open source code freely available on Github that Matz says anyone can use to measure gene expression on a whole-genome scale. "They are low cost, super efficient, and applicable to non-model organisms," Matz said. "Now, anybody can take their own favorite worm, bird, fly, or flower, and do basically the same kind of experiments applying the same methodology as we developed and showcased in this paper."
Reef conservationists could take adult corals from a warm location and transplant them to vulnerable regions and let them naturally breed within local populations. "You're just helping the natural process of mutations spread through coral populations. This is a very important message for reef managers. People are now seriously considering switching their reef restoration programs to achieve these results. It's pretty cool. I'm pretty happy about this," Matz said.
The study, "Genomic determinants of coral heat tolerance across latitudes," appeared in the journal Science, June of 2015. The authors are Groves Dixon, Sarah Davies, Galina Aglyamova, and Mikhail Matz of the Department of Integrative Biology, University of Texas at Austin; Eli Meyer, Department of Integrative Biology, Oregon State University; and Line Bay, Australian Institute of Marine Science. Study funding came from the National Science Foundation and the Australian Institute of Marine Science. Bioinformatics analysis was performed with computational resources of the Texas Advanced Computer Center, The University of Texas at Austin.
