Scientific Computing

Preventing Freezer Burn and Saving Human Tissue

It sounds simple, but Shah Razul’s work involves trying to understand what water molecules do when they get to below zero degrees Celsius because the possibilities that understanding offers could be life-changing.

When water crystallizes, its volume increases and it forms a regular structure that destroys, for example, cells, or, in the case of an organism, its delicate membranes and systems. As most of us will know, when food is frozen and we leave it too long, the ice separates out, leaving the food “freezer burned.” A wasted chicken breast is one thing, but this problem also exists, for example, in cryopreservation of cells, tissues and organs.

“When water’s in a close-to-frozen state, it has some interesting behaviour,” says Razul, an associate professor of chemistry at St. Francis Xavier University. “We’re trying to understand what the water molecules are doing.”

Razul’s overarching goal is to find a way to keep water from forming crystals and then applying that learning to real-world problems in areas such as food and health care.

“If we can, we can solve many problems related to freezing,” he says, including, for example, potentially creating environmentally safer antifreeze products to replace or minimize the use of road salt.

He looks to creatures such as the wood frog, which uses glucose to stop the freezing process from destroying its tissues in winter. Razul, therefore, is running simulations to look at the process of freezing and analyze what water molecules are doing in the fractions of a second just before they form crystals and freeze.

“A lot of my work now is involved with understanding how small biomolecules, such as different kinds of sugars and salts, stop water from freezing or slow down the process,” he says, adding that he wants to figure out a biomolecular combination that would achieve this goal.

Over the past five years, he’s been using computational principles to develop a cryoprotectant to preserve lobster meat.

“We’ve tried it and it works,” he says, adding that some companies from overseas are testing his system currently. “We had a taste test in Atlantic Canada, and we’ve published a couple of studies where the public indicated that it tastes as if it was cooked yesterday, preferring it to frozen lobster meat that is sold currently.”

His next frontier is looking at the ways in which his process can preserve muscle cells and brain cells to see what kind of applications it might have in the healthcare field.

“No one has a definitive answer to how these cryoprotectants work at the molecular level,” he says. “It’s basically all trial and error. We try a little bit of this and a little bit of that and see whether it works.”

To do his work, he’s watching the behaviour of water in minute detail — fractions of seconds — and then he replicates the test multiple times.

“Because of that, I generate a lot of data,” he says, and that’s why he couldn’t do his work without ACENET’s high performance computing.

“It’s almost impossible to do this work without ACENET,” he says. “I would need a year and a half to do it sequentially as opposed to one month with ACENET.”

Putting the Stars in Perspective

Catherine Lovekin’s main research is dedicated to understanding the heat-transfer process in stars. In particular, she studies the “convective overshoot” where the circular currents that exist in stars go a little beyond their theoretical boundaries.

“We don’t really know how much of this overshoot there should be, so I use asteroseismology, which is a way of measuring the variations in stars, to try to figure that out,” explains Lovekin, associate professor of physics at New Brunswick’s Mount Allison University. “Stars pulsate and they have variations like waves on a head of a drum or something. And I’m trying to use those waves to understand the interior of the stars.”

The overshoot is important, she says, as it changes how big the core of the star is, so it also changes all kinds of things about the subsequent evolution, and what elements are produced — for example, how much fuel it has for nuclear fusion. It can change the lifetime of the star.

The hope is that eventually she and others who are working on the same question will get enough of a sample that they can start to see patterns. For example, maybe the overshoot behaves differently in high mass stars compared to low mass stars, or maybe it changes as stars evolve, meaning that older stars have different values than younger stars.

While Lovekin is working mostly on individual stars, her students are looking at stars that have pulsations and that are part of a binary system — in effect, two stars in orbit around each other. Looking at the stars this way, she says, is a good way to constrain things such as the diameter, temperature and mass of the star.

As her research has expanded to the way in which binary stars behave, her models have become that much more computationally intensive, to the point where she now can’t do her work without the services of high-performance computing (HPC) — in this case, from ACENET.

“I could do some aspects of my work without those resources, but it would take me years longer without something like ACENET.”

Lovekin’s relationship with HPC goes back a long way. She was an early user of the HPC computing from the Digital Research Alliance of Canada, then called Compute Canada, which is the national partner of ACENET. And today, her students are all using ACENET.

“The ACENET training that is offered every May is so useful,” she says. “I just tell my students to sign up. It really helps get them started. The ACENET folks have so much more experience with helping to train people on the system. When my students start working with me, they all know how to run these models because of their ACENET training.”

Building Better Brain Imaging Analysis Tools

Erin Mazerolle uses brain imaging to understand, for example, what’s happening in the brain of someone with Alzheimer’s disease.

“A lot of the work that I have done is to try to do a better job of understanding and diagnosing neurological diseases, and to take really advanced cutting-edge research imaging techniques — new ways to image the brain — and apply them for the first time in patient populations,” says Mazerolle, assistant professor of psychology and computer science at St. Francis Xavier University.

Over the course of this research, Mazerolle noticed that she and her peers don’t always do a good job at making sure their results are reproducible in other studies.

“If I do a study and publish it, and then another researcher reads it, can they get the same result?” she asks. “This is the part of my work where ACENET really comes in. I have a study going on right now in which we’re trying to figure out how the different ways that we analyze data are impacting the reproducibility of our results.”

She says that when analyzing brain imaging data, there are dozens of steps researchers take, and each one involves deciding which parameter to use at that juncture. And ultimately, all of those decisions impact the final result.

“There’s an analogy that we use called the ‘garden of forking paths.’ I make a decision and go down one garden path and I never know what my results would have looked like over on the other path,” she says.

“So we’re developing software that researchers can use to simultaneously analyze the imaging data using a whole bunch of different paths. That’s a really major computational challenge because these datasets are very large and it takes a lot of computational time and power to get to the end of each of these paths.”

She says the clusters at ACENET reduce the computational challenge to a reasonable time frame, but even with those high-performance computing resources, it can still take a month to do the analysis for a single study.

“It would be completely unfeasible for me to do it without ACENET,” Mazerolle says.

“We hope that by looking at all of the outcomes of the different paths, we get a better sense of whether our results are real, or if the ones that we like only occur down one or two paths,” she says. “Or, do the 200 paths actually look totally different and we never would have known that if we hadn’t tried this multiverse analysis?”

For her multiverse research, she’s using data from the Alzheimer’s Disease Neuroimaging Initiative, an international data-sharing database that looks at Alzheimer’s Disease, cognitive impairment and healthy normal aging. In particular, she’s comparing super-agers (who are in their 80s or 90s and have the memories of 20- and 30-year-olds) and comparing them to people who are aging normally. This is the study that will pilot the tool she hopes will help further brain imaging research across Canada.

Mining Data to Mine Metals, Among Other Things

Geophysics professor has the key to determining how deep down a mineral deposit might be without digging a hole.

Geophysicist Colin Farquharson is a mining executive’s dream come true and maybe even a godsend to a detectorist with deep pockets.

The professor at Memorial University’s Department of Earth Sciences can take the data from geophysical tools that collect information on everything from metallic ore deposits to water sources and determine how deep these prospects are, unlocking great potential for whomever has asked him to do so — or at least information on whether they should pursue a specific site.

“Any scenario where you want to figure out what’s down in the ground without digging a hole — that’s where geophysics comes in,” Farquharson explains.

The equipment used by prospectors is often a helicopter that slings out a loop-shaped transmitter that’s 20 metres in diameter. The transmitter loop has electric current running through it and will offer a response if the helicopter happens to fly over, for example, an ore deposit, even if it’s embedded in rock.

“Metallic ore deposits tend to have a higher electrical conductivity than the rocks they’re sitting in, so they can conduct electricity, which is a bit surprising but true,” he says. “So these are geophysical methods where it’s exactly like the metal detectors people use when they’re trying to find relics and treasure.”

The main application for what Farquharson does is mining, but there are others, too. Often, instead of precious metals, geophysical equipment is used to look for sources of water, particularly in developing countries that might have fresh water aquifers near the coast with ocean water causing problems by creeping in.

Once the equipment determines there is something desirable there, the next thing the prospectors want to know is how hard it’ll be to unearth.

“People want to know how deep those deposits might be to determine whether it’s worthwhile pursuing,” Farquharson says. “That’s where computer modelling comes in. We use the information to calculate synthetic data.”

But making those calculations can be expensive and computationally challenging so Farquharson takes advantage of the high-performance computing power offered by ACENET.

“We can do in hours what would otherwise take weeks and weeks,” says Farquharson, who often gets his graduate students to run the datasets. “Some of the datasets are rich. Even on ACENET some of my students find it’s taking a few days to do these computations.”

He says the support services ACENET provides have been extremely useful to him and his students.

“It’s really valuable to me to have ACENET and the Digital Research Alliance of Canada, which both have big computers that are accessible for free. It doesn’t cost me anything for what I do, so it means I’m not spending thousands of dollars to do that part of my work.”

Probing the Infinitesimally Small with Laser Pulses

Samira Barmaki develops numerical experiments to investigate the effect of intense and ultrashort laser pulses on matter.

“We are currently witnessing significant experimental progress in laser pulses,” says Barmaki, a physics professor at the Université de Moncton’s Shippagan campus. “The durations of these pulses are becoming increasingly short, with some on the order of 100s of attoseconds or less [an attosecond is equal to one quintillionth of a second.] These shortened pulses are excellent tools to help us probe and control the movement of electrons in atoms and molecules — the properties of matter.”

Barmaki and her team at the computational and photonic physics laboratory create simulations to help find ways to control the way electrons leave atoms and the amount of energy they will share. Her work uses numerical methods to accurately describe the energy spectrum of simple atomic and ionic systems, and her group has recently been studying “doubly excited” electronic states that form in the atomic energy spectrum following an excitation by an XUV laser pulse.

“This investigation has made it possible to detect and characterize new doubly excited states [DES] never observed to date,” Barmaki says. “The results of the simulations we are developing will also serve as support for future experimental studies, whether for calibrating laser parameters, proposing new investigation techniques or helping to analyze and explain the resonant features that manifest themselves in the recorded ionization signal due to the involvement of the DES.”

Barmaki says she couldn’t do her work without high-performance computing, which she uses to develop complex algorithms “capable of describing with very high resolution the interaction of the targeted atom with the laser pulse.

“The high-performance computing used in our experiments is based on developed theories using quantum physics, atomic and molecular physics and photonics,” she adds.

Developing these algorithms requires powerful computing and technical resources such as those provided by ACENET and the Digital Research Alliance of Canada.

“Such data processing powers allow us to carry out numerous large-scale projects at our lab and an essential part of that work involves the training and supervision of students at all levels,” she says. “Whether undergraduates for summer internships, master’s students or doctoral students, they are all trained in computational physics and the exploitation of existing algorithms, and they develop their own algorithms by having access to ACENET and the Alliance’s supercomputers.”

Using Modelling to Make Energy Use More Efficient

Kush Bubbar came to academia from industry. Leveraging his extensive background in technology development, including semiconductors, biomedical engineering, telecommunications, manufacturing, and renewable energy, Bubbar leads the Sys-MoDEL lab within the Faculty of Engineering at the University of New Brunswick, where he focuses on delivering value to societal projects by applying methods at the forefront of innovation into his academic practice.

The lab’s mission is to understand complex systems and present their clients with feasible, yet optimal solutions to address their challenges.

The lab’s diverse project portfolio includes advanced vehicle dynamics, marine renewable energy, and renewable energy system optimization. As Bubbar explains, “Our work on integrating renewable energy sources into existing power systems and optimizing wave energy conversion processes requires substantial computational power, often at a scale that would be unmanageable without the resources ACENET provides.”

Currently, Sys-MoDEL is engaged in three pivotal projects – power system planning with NB Power, oceanic wave energy conversion with Sapphire Energy, and off-road vehicle design with Potential Motors.

For the NB Power project, “We’re trying to understand how to improve the planning process for power systems in the future with the knowledge that we are integrating more renewables in there,” he says. “We are looking into how we can incorporate energy storage as a means to supplement our transmission network.”

Bubbar says this project is very novel, but also very computationally heavy.

“Under normal operating conditions, it’s easy to understand, but it’s at those times and under the conditions where you have faults, reliability issues, when things break, that you want to ensure the system is as reliable as it can be,” he says, referring, for example, to a wind storm across the province. “We’re looking at these cases where there is huge opportunity to reduce cost, and to be more efficient, through implementing energy storage mechanisms into the transmission system.”

“These are computationally intensive tasks that require robust simulation capabilities. ACENET’s supercomputing systems enable us to perform these large-scale simulations efficiently, reducing both time and cost,” Bubbar states.

ACENET not only provides the necessary computational infrastructure but also offers a supportive ecosystem for research development. This includes training modules, expert support, and a network of resources that are instrumental in troubleshooting and refining research methodologies. “The value of ACENET extends beyond just hardware. Their training and support have been crucial in helping us set up and optimize our computational experiments,” Bubbar says. “If my students have trouble for any reason, they can connect with a local resource on the UNB campus.”

By collaborating with ACENET, Sys-MoDEL gains a strategic advantage, ensuring that Bubbar and his team can continue to push the boundaries of what is possible in researching complex systems. This partnership exemplifies the synergy between advanced computing resources and innovative research, highlighting ACENET’s role in accelerating scientific discovery and technological development across the Atlantic region.

Record Use for UPEI!

Physicist James Polson wanted to tackle a computational project inspired by a collection of experiments done a few years ago by a research group at McGill University. He knew the project would be computationally demanding, but he had no idea that by the end of it, Matthew Kozma, his research student, would have set a Canadian record for compute-cycle use among undergraduate students. When you consider that most high-performance computing work is done by graduate students, postdocs and faculty, it’s noteworthy that an undergraduate — who only worked in the summer months — from a very small Atlantic Canadian university used an amount on par with Canada’s most compute-intensive researchers.

The use was immediately evident to ACENET. “Suddenly a single research group was using more compute resources than entire provinces,” says Greg Lukeman, ACENET CEO.

Polson’s project started with data from the Walter Reisner research group in the Physics Department at McGill University, which does experiments that examine the physical properties of DNA.

“It’s for the purpose of advancing a certain type of nanotechnology designed to manipulate and analyze DNA molecules,” explains Polson, a professor of physics at UPEI. “So they’ve devised this experiment to look at how DNA molecules behave when you squeeze them into very small and heterogeneous spaces.”

Polson had wanted to do a computer simulation project to complement their experiments. He knew what he wanted to do, but the computational methodology was new to him.

“It was on my to-do list for a number of years, but I was always reluctant to give it to an undergraduate student so I figured it would be something I’d have to do myself,” Polson says. That is, until Matt Kozma arrived at UPEI. “Getting the methodology implemented so that you can do these calculations is a really difficult thing. But Matt is very computer savvy, so I thought, ‘if there’s any undergraduate student who can tackle this project, it’s Matthew.’”

Kozma wrote the computer code from scratch, which “is not a trivial thing to do,” according to Polson. “And he was also able to implement the methodology.”

They set out the methodology and code over the course of the summer of 2021, and made sure the program they’d devised was well tested. Then, in the summer of 2022, Kozma really got going. Using all five of Canada’s national supercomputers, his simulations totalled over 30 million CPU hours, setting the all-time Canadian record for usage by an undergraduate student, and more than doubling the previous record. In fact, during his five-month run, he was the third highest user (undergraduate or otherwise) in Canada. Kozma worked elsewhere in the summer of 2023, but when he returned to his studies this year, he and Polson finished up the last bits of work on the study and their paper will be published later this year.

Kozma presented the research at the Atlantic Universities Physics and Astronomy Conference two summers in a row, as well as at the Canadian Undergraduate Physics Conference.

“It’s always great to have the opportunity to present the research, in addition to doing it,” Kozma says.

Ultimately, their research has applications “way downstream” in advancing nanotechnology for manipulating and analyzing DNA molecules.

“Our calculations should help contribute to the improvement in the design of nano-devices for biomolecule analysis,” Polson says. “And yes, far far downstream, this will perhaps lead to important benefits in the health sciences.”

In the meantime, the experimentalists at McGill have expressed considerable interest in the calculations carried out by Kozma and Polson.

Polson says he hopes this story will illuminate the fact that undergraduate researchers can contribute a lot when the right supervision is in place, and when they have access to the right resources.

What Drives Galaxy Evolution?

Dr. Ivana Damjanov studies the relationship between galaxy size growth, changes in galaxies’ stellar and dark matter content and their position in the universe.

“We are interested in the connection between the evolution of galaxies and their surroundings as cosmic time goes on,” says Damjanov, associate professor at St. Mary’s University. “Galaxies can live either surrounded by basically nothing or they can be in very dense regions of the universe called galaxy clusters.”

She and her team zoom in on galaxies to see how they change their appearance in telescope images.

“We take the images of millions of galaxies and we quantify their appearance,” she says. “We measure their size, shape, how squished they are and we compare that to how many stars they’re making, how fast these stars are moving and what they are surrounded by. Our goal is to understand how the changes galaxies are going through work together and which physical processes drive those changes.”

While answering these fundamental questions, Damjanov and her team are also developing techniques that can be used in everyday life. The techniques she’s using to analyze her large-scale images are advanced statistical methods and the instruments they’ve been developing are used, for example, in medical imaging.

Dr. Michele Pizzardo, Damjanov’s postdoctoral fellow, is using ACENET to see how simulated clusters of galaxies evolve and grow.

“We are interested in the study of these galaxy clusters,” Pizzardo says. “I’m interested in the external part of these structures. This part is sensitive to the basic laws the universe follows. It can give us a way to test different cosmological models to determine whether a model is viable.”

Using ACENET, he developed a “recipe” to understand how much these clusters are accreting at different ages of the universe.

“We need a lot of clusters and then we use all the information — the temperature, entropy, energy information, position of each particle in these boxes,” he says. “We do this at different ages of the universe. One dataset can be 40 to 50 terabytes of data.”

Obviously, he had to use ACENET’s high performance computing technology to do these calculations. Damjanov adds that other members of her team also use ACENET to analyze actual images from telescopes.

“In principle, our software could be run on a standard computer,” Pizzardo says. “ However, the size of the data that our software processes makes it impossible to obtain any results. In other words, if the data we had to analyze were a few GBs, we could run the software on a local machine, but our data is more than 1,000 times larger than that so the use of a cluster infrastructure is essential.”

He says the team members’ models are computationally demanding because their random access memory needs can climb up to 100 gigabytes and they need dozens of high-performance nodes to obtain results in a reasonable amount of time. Storage needs are another consideration.

In other projects, Damjanov’s students are creating simulated galaxies and adding them to the existing telescope images, thus bringing together voluminous simulated and real-world datasets. ACENET is essential for that big-data-driven work.

Simulating Solutions: The Power of Computational Chemistry

Stijn De Baerdemacker says his research in theoretical chemistry is “fairly fundamental” and yet it doesn’t find itself very far away from real-world applications.

“Typically, when we think about chemistry, it’s about tubes and beakers and people in lab coats, and that’s still a big portion of what of what we do,” De Baerdemacker says. “However, it’s not only running the experiments, it’s also exploring what molecules you can make and how they can solve problems.”

In other words, he’s always looking to develop something new — a better material for a device, or a better drug for a disease, for example. And if you’re searching for the “holy grail” that will solve your problem, it’s very time-consuming, he says, because you have to assess all of the options.

“This is where computations come in, because on a computer, we are not bound by safety requirements,” he says. “We can just go in and try to simulate what will happen. That speeds up the discovery process by many orders of magnitude.”

De Baerdemacker applies various mathematical models that describe the structure of molecules to determine which ones would provide the most accurate results.

“We make sure the methods are grounded in proper theory,” he says. “Once we have the theory down, we run tests and we use ACENET’s resources to put everything into code. We test it on our own local systems and then try them on a bigger system like ACENET,” he says. They then send the code to the chemist, who runs computer simulations using various molecules. This enables the chemist to narrow the options to the most promising candidates before running experiments.

De Baerdemacker says that when he talks to people about high-performance computing and asks them who they think might be HPC’s biggest clients, they’re surprised to learn that chemists are among them. And, he adds, his colleagues in computational chemistry are definitely more frequent users, but even fundamental chemists make good use of services at ACENET.

“We also run a lot of simulations,” he says.

Using a concrete example, he says a lot of diseases are associated with how the proteins in our bodies work and the way they are folded in the cell is also are important for their functionality. Alzheimer’s is one example where a protein starts to curl up and then pierces through a cell membrane,” he says.

“We’re actively looking for a drug that will inhibit that behaviour, but in order to do that, we need to come up with candidates and do a lot of simulations,” he says. “When we’re doing those simulations, we use ACENET.”

But where he has used ACENET to the fullest extent is with machine learning, which he says chemists have been using for 20 years.

“When we’re working on computations, it’s not one variable, it’s millions,” he says. “On machine learning, we’ve discovered that the machine has figured out the kind of molecule we were looking at all by itself. It looked at the data and it found patterns. At this point, ACENET is really crucial.”

A Prize-Winning User

Erin Johnson wouldn’t have won the prestigious Steacie Prize, awarded to early-career scientists, in 2021, without the Digital Research Alliance of Canada’s high-performance computing and the services of ACENET. “That’s a completely fair statement,” says Johnson, who was an early adopter of high-performance computing, dating back to her student days at Carleton University, Queen’s University and Dalhousie University, as well as her post-doctoral work at Duke University. She’s now a professor and Herzberg-Becke chair in theoretical chemistry at Dalhousie University. Asked how she would explain her job to a guest at a cocktail party, Johnson says she studies intermolecular interactions within the materials all around us. “These are the weaker interactions between molecules as opposed to the stronger bonds within a molecule,” Johnson explains. “So, we would, for example, at a cocktail party look at the interactions in beer or wine in a glass — those types of interactions within a liquid or a gas or within a molecular crystal.” She does this to predict such properties as reactivity, hardness, and conductivity, among others. “Any chemical observable is something that we can predict through this type of modelling,” she says. Johnson’s lab, however, might surprise those who envision her working in a chemistry lab full of beakers and bunsen burners. Rather, it’s a computer lab from which she’s developed methods for modeling that are some of the most accurate and efficient available. It was those methods that won her the Steacie Prize. “We then take those methods and apply them to problems in chemistry,” she explains. “One of the particular problems we’re focusing on is the problem of molecular crystal structure prediction or how molecules would come together to form a 3D solid.” She likened the ways molecules come together to the ways in which Lego bricks do. “There are many ways you could conceive of molecules coming and packing together, but not all of those are going to be stable,” she says. “So the challenge is trying to predict how they will actually pack in a solid.” Once you solve that puzzle, there are applications across several industries, one of the most recognizable being pharmaceuticals. “If you are producing a drug and you want it in pill form, you want a solid but soluble state,” Johnson says. “When you try to formulate new drugs, you want to screen for all the possible polymorphs (transformations to another form) and find out which ones are the most stable. What you don’t want is for it to easily form a particular polymorph that’s not a very stable one and then change over time so that it’s no longer soluble.” These methods also have applications for electronics, but instead of looking at solubility, the property of interest would be conductivity. “Like the solubility, the ability of charge to flow through a material is also dependent on the particular polymorph” Johnson explains. “You can think of many applications where the different solid state properties between polymorphs would affect whether a material was promising or not.” Johnson says her work would be “completely impossible” without ACENET’s services.