Supersized Science

This podcast features an interview with biologist Mikhail Matz, Department of Integrative Biology, College of Natural Sciences, The University of Texas at Austin. Matz was part of a study funded by the National Science Foundation and the Australian Institute of Marine Science. In June of 2015 they published results in the journal Science that found the first evidence that corals can genetically adapt to warmer waters from climate change. Podast host Jorge Salazar interviewed Matz about his computationally-based findings and about open source tools other scientists can freely use to analyze genomes of plants and animals.

Podcast host Jorge Salazar interviews scientists Michael Sacks, Institute for Computational Engineering and Sciences at the University of Texas at Austin; and Ming-Chen Hsu, Department of Mechanical Engineering at Iowa State University. New supercomputer simulations have come closer than ever to capturing real behavior of human heart valves. The studies focused on how heart valve tissue realistically responds to blood flow. And to be clear this is ongoing research, meaning they don't have all the answers yet, but they do say they've made progress on a really tough problem that potentially affects hundreds of thousands of people each year with hearth disease. The scientists say their new supercomputer models can potentially help doctors make more durable repair and replacement of heart valves.

Host Jorge Salazar interviews scientists Karen Vasquez and Albino Bacolla of the University of Texas at Austin. Supercomputers have helped scientists find a surprising link between cross-shaped pieces of DNA and human cancer, according to a study at The University of Texas at Austin. DNA naturally folds itself into cross-shaped structures called cruciforms that jut out along the sprawling length of its double helix. The DNA cruciforms typically aren't anything to worry about. In fact, previous evidence show that DNA cruciforms are essential to life. They enable DNA replication, part of how cells make copies of themselves. And they help initiate gene expression, which makes proteins. What's more small DNA cruciforms are commonly found inside our bodies. Scientists estimate as many as 500,000 cruciform-forming sequences of DNA can exist on average in a normal human genome. What the UT scientists are doing is investigating the origins of human cancer. And what they've found is that these tiny cruciforms - just a small shape of normal DNA - are linked to mutations that can elevate cancer risk.  

Host Jorge Salazar interviews scientists Min Chen of Rice University and Jeroen Tromp of Princeton University. An international science team reported a discovery of gigantic rock structures hidden deep under East Asia, centered on the Tibetan Plateau. Scientists used supercomputers to process earthquake data and make images in 3-D down to depths of about 900 kilometers, or about 560 miles below ground. Scientists from China, Canada, and the U.S. worked together to publish their results March of 2015 in the American Geophysical Union Journal of Geophysical Research, Solid Earth. The study area is a hotspot for earthquakes. And it's surrounded by networks of seismographic stations, 1869 stations in all. That's where scientists got their data to take cat scans of the Earth using the supercomputer model they developed. The science team says their research could potentially help discover hidden pockets of hydrocarbon resources like oil and gas. More broadly they say their work will help explore the Earth hidden miles under East Asia and elsewhere.

Host Jorge Salazar reports from the Texas Advanced Computing Center an interview with Michael Grabe, an associate professor in the Department of Pharmaceutical Chemistry and the Cardiovascular Research Institute at the University of California, San Francisco. For the first time ever, scientists designed completely from scratch a protein molecule that behaves like a slice of life. It mimics a natural protein found in living cells that transports ions across a cell membrane. The cell membrane surrounds living   cells like an envelope. And ion transport through the membrane helps keep us alive. It lets nutrients in and waste out of cells, and it also transmits signals between nerve cells of the brain and spinal cord. Scientists used the Stampede supercomputer at TACC to model the stability and dynamics of the designed protein. They did this with an allocation through XSEDE, the Extreme Science and Engineering Discovery Environment, funded by the National Science Foundation. The researchers published their results in the journal Science in December 2014.  This research has wide potential application, such as targeting medicines more specifically into cancer cells and driving charge separation potentially for harvesting energy for batteries.

Matthew Hanlon manages the Web and Mobile Applications Group at the Texas Advanced Computing Center. And Matt Vaughn directs the Life Sciences Computing Group at TACC. Vaughn and Hanlon present a one-hour core conversation for South by Southwest Interactive on Monday, March 16. It's called A Next Generation Platform for Open Data. In the podcast they discuss their work on the Arabidopsis Information Portal, a new online resource for plant biology research. Matthew Hanlon: We're looking to attract both data scientists and portal developers, anyone who has experienced developing, hosting, running or trying to market an open data portal to the community. Matthew Vaughn: The Arabidopsis Information Portal, Araport for short, serves two purposes. It's a clearinghouse for genetic, genomic, protein and gene expression information for the model plant Arabidopsis thaliana… But it's also a resource for people who build portals.

John Fonner is part of the Life Sciences Computing Group at TACC, and Rion Dooley manages the web and cloud services group plus he leads the Agave API project. Fonner and Dooley present a hands-on four-hour workshop for South by Southwest Interactive on Tuesday, March 17th. It's called Science as a Service and other Lies: Hacking Big Data on the Web. In the podcast they discuss science-as-a-service. Rion Dooley: When we talk about science as a service, what we're really talking about is the ability to think about your science and interact with your data and with your experiments in a way that is thinking in terms of the problems you're trying to solve rather than the machines you're trying to solve them on. John Fonner: The domain field is a patchwork of individual codes out there that have very specific assumptions about the type of system that you're running them on. The types of algorithms might have very different hardware requirements. And the data volumes now for a lot of the next-generation sequencing questions are just huge. We want to talk about the tools that we have and the way that we've been tackling this problem and really taking it from a place that is inaccessible to most biologists and only accessible to the most technically savy scientists, and putting it in a place where pretty much anyone can have access to it.

Maytal Dahan is a research engineer and science associate at the Texas Advanced Computing Center. She presents a meet up for South by Southwest Interactive on Saturday, March 14. It's called Science and Tech Meet Up: Scientists and Geeks Unite!. In the podcast Dahan talks about some interesting intersections where science meets software development, and she predicts the next big thing at SXSW -- science! Maytal Dahan: It'd be really great if we could get developers of all kinds – user interface, web, application developers – who are interested to learn how science impacts the world we live in today, and how to engage and possibly contribute their skills to science projects…A developer at a start-up doesn't always get to meet a scientist creating an API for their application.

Host Jorge Salazar reports from the Texas Advanced Computing Center an interview with Klaus Schulten, professor of Physics at the University of Illinois at Urbana-Champaign.One of life's strongest bonds has been discovered by a science team doing research on biofuels with the help of supercomputers. The biomolecular interaction binds at about half the strength of a chemical covalent bond the pieces of a finger-like system of proteins called cellulosomes used by bacteria in cow stomachs to digest plants. The researchers published their results in the journal Nature Communications in December of 2014. Their find could boost efforts to develop catalysts for biofuel production from non-food waste plants.

Materials with novel electrical properties discovered using XSEDE computational resources Stampede and Lonestar supercomputers of TACC. This podcast features an interview with materials research scientist Xiaofeng Qian of Texas A&M University.

Scientists have used supercomputers to find what they say is the best evidence yet that a plant's genes  sensitive to cold and drought will help it adapt to changes in its environment. What they studied in plants was gene expression, instructions coded in DNA that regulate how many proteins it makes. Gene expression gives rise to traits such as tolerance of cold or drought, and it can evolve through natural selection to help a plant cope with environments out of its comfort zone.  The results of the computational biology study were published in the journal Molecular Biology and Evolution in September of 2014. In it scientists studied the flowering mustard weed Arabidopsis thaliana, known as a model plant in part because it has one of the smallest genomes, which was completely sequenced in 2000. The science team first took Arabidopsis genes found in the lab from a prior study that respond to cold and drought. They then compared those to reference genomic data from over a thousand strains collected throughout Europe and Asia.  Finding associations was like finding a needle in a haystack, and to do that they enlisted the help of the iPlant collaborative and they used the Ranger and Lonestar supercomputers of the Texas Advanced Computing Center.The study was co-authored by Thomas Juenger, a faculty member in the Department of Integrative Biology of the University of Texas at Austin.

Astrophysicists have been puzzled by their observations since the 1970s that only a small fraction of matter in the cloud becomes a star and part of a galaxy. They found a lot less of the universe's mass than expected in the middle of galaxies. Things changed when a multi-university collaboration produced a set of new supercomputer models of galaxies called FIRE, The Feedback in Realistic Environments. FIRE simulations ran on the Stampede supercomputer at TACC, an XSEDE resource funded by the National Science Foundation. Theoretical astrophysicist Philip Hopkins of CalTech led a 2014 study of initial results that found that star activity - like supernova explosions or even just starlight - plays a big part in the formation of other stars and the growth of galaxies. Philip Hopkins spoke more about galaxies on FIRE.

Sathoshi Matsuoka came to the supercomputing conference SC14 and received the 2014 IEEE Computer Society Sidney Fernbach Award for innovation in the application of high performance computers. Dr. Matsuoka is a professor and leader of the Tsubame project, one of the world's fastest and most efficient supercomputing grid clusters. Tsubame is at the Global Scientific Information and Computing Center, hosted by the Tokyo Institute of Technology. We spoke with Sathoshi Matsuoka on the opening night of a busy convention floor at SC14.

Thomas Sterling spoke on a couple of panels at the supercomputing conference SC14 that looked at the promises and pitfalls on the path to developing exascale supercomputers, the next-generation of the world's fastest computers. Dr. Sterling is the executive associate director and chief scientist at the Center for Research in Extreme Scale Technologies at Indiana University, Bloomington. He's won the Gordon Bell Prize for innovations in high performance computing, and he laid the foundation for the current paradigm of supercomputers by co-developing the Beowulf cluster of commodity Linux cluster computing. Sterling's latest project is the ParalleX execution model being tested in part on XSEDE resources here at TACC with the Stampede supercomputer.

Best-selling author and physicist Brian Greene of Columbia University gave the keynote address at the SC14 conference. His latest book is The Hidden Reality: Parallel Universes and the Deep Laws of the Cosmos. In it Greene describes the hypothesis of multiple universes, and in particular that of a computer simulated multiverse. Brian Greene spoke with me by phone about the possibilities and future of supercomputing.