Each day, across the globe, dedicated scientists from all disciplines pit themselves against humanity’s most perplexing problems. Between them and potential solutions stand puzzles so complicated and difficult to address — reverse-engineering the brain, providing energy from fusion, limiting drug resistance, for example — that they require a staggering amount of resources. Some riddles, like mapping the human genome, have been solved. Others remain defiant. Scientists call these mysteries Grand Challenges. 

A relatively new field of research called computational science is helping researchers gain ground in battles that have, to this point, favored mystery over discovery. Bob Gotwals, Instructor of Chemistry at NCSSM and pioneer of the school’s computational science program, has been championing the field at the school since he came to NCSSM as a full-time faculty member in 2006. Now NCSSM is unique among high schools for its expertise in the field and the extensive computational science courses it offers.

Put simply, computational scientists use computers to do much of what experimental scientists do through more traditional means — say, beakers and burners. It seems an unlikely scenario until you understand that science, in the most traditional terms, is the manipulation of known entities, with known properties, to generate new entities or a deeper understanding of existing ones. These known properties can be defined numerically; Entity A has a certain unique property (let’s assign it the number 1), Entity B has its own unique property (let’s call it 2), and when these entities are put together, the result is Entity C which, as you would expect, has a numerical property equivalent to 3. In other words, 1+2=3. Easy enough for basic problem sets.

But imagine an experiment where multiple complex entities — each with a fingerprint hundreds of numeral characters in length — need to be combined. Finding the magic combination that results in a cure for Alzheimer’s disease, or a genetically engineered plant that eliminates malnourishment in undeveloped countries, may require tweaking every number in the fingerprint of every entity in the experiment down to the one-thousandth or the one-millionth, over and over and over.

It’s a near-impossible task — unless a computer does it for you. 

With such digital horsepower available at the stroke of a key, computational scientists can employ mathematical modeling, algorithms, and sheer repetition to simulate and analyze highly complex problems, often in a fraction of the time it would require through traditional trial and error in a laboratory. Modifying the “experiment” is a matter of adjusting the numbers and hitting “run.” With a bewildering amount of data coursing through circuitry nearly at the speed of light, weeks or months or years of research can be reduced to minutes, hours, or days.

Gotwals says the technique can avoid wasted time and expense by spotting an experiment doomed to failure. “I can run that model computationally,” he says, and at the end, “I can tell the scientist, ‘You can spend the next 20 years trying to get this reaction to run, [but] it’s not going to run.’”

Not an either/or

As the new kid on the block, it was inevitable that computational science would meet with pockets of mild resistance, particularly in the earlier years of the field’s development when its true power was not yet fully understood. “There’s something to that” initial distrust of the field, says Charlie Payne, who teaches computational physics at NCSSM. Payne, who has incorporated elements of computational thinking in his physics courses throughout his career, has been programming since his student days in the mid-1960s. “In those days, we didn’t have the same data that we do now, but as time has gone along, at least in my field, you see a lot more collaboration.”

Gotwals admits to having been branded by other chemists as not a real chemist. “It’s still chemistry,” Gotwals counters, “just without the smoke and fire.”

By no stretch of the imagination, however, has computational science replaced hands-on science in the laboratory, says Mariah King, a computational chemist teaching at NCSSM. Instead, she says, it’s one of a “triad” of complementary fields — theoretical, experimental, and computational.

The computational approach, King says, “exposes everyone to more information. It’s the idea of looking at one question from 10 different perspectives. It’s why we talk now about this interdisciplinary approach, because the chemist’s approach isn’t perfect; the physicist’s approach isn’t perfect; the biologist’s approach isn’t perfect, or the mathematician’s.”

And neither is the computational scientist’s. Is it faster? Exponentially. But their results are just as conditional as those from more traditional approaches.

“A number of assumptions are still being made,” King says. “Just because we’re using algorithms that are able to generate data very quickly doesn’t mean that all that data is a hundred percent accurate. At the end you have to explain to the experimentalist, ‘Well, these are the assumptions that went into this model and so you have to keep those assumptions in mind.’ ”

And no matter the computational outcome, the claims made by computational scientists still need to be verified by scientists in the lab, says Payne. “An experimentalist has to say, ‘Yeah, it works,’ or ‘No.’ ” 

“Altogether,” King says, “you get to a bigger picture of how to approach that single problem.”

Computational science at NCSSM

As far as Gotwals and his colleagues are aware, NCSSM is the only high school in the country that offers a credit-bearing suite of courses in computational science. Interested students can choose from 11 courses that range from an introduction to the science to computational investigations of physics, bioinformatics, nanotechnology, and medicinal chemistry. Nine of the 11 courses are taught through the school’s Online program, while two are offered residentially. 

Emma “Em” Ambrosius, a senior at Atkins High School in Winston-Salem and veteran of the NCSSM Online program, is one of the relatively few students in the country who have been able to take computational science courses while still in high school. This semester she is taking two online courses with King: Industrial Chemistry and Chemical Engineering, and Computational Chemistry.

“You know you’re having opportunities and you’re able to put things on your resume that no one else is,” Em says of her enrollment in the courses. “I’m trying to market myself and be the best [college] applicant I can, and it’s really helpful in that.”

NCSSM and its computational science students owe this unique opportunity to two things: First, says King, is the incredible luxury of having access to necessary resources not always available to more traditional public schools. The second, says Gotwals, is the foresight of retired science dean and faculty emerita Myra Halpin.

“The woman’s a visionary,” says Gotwals of Halpin, who brought him to NCSSM specifically to build the computational science program. “She recognized that we were not going to be competitive as a school if we didn’t get on the computational bandwagon. She’s an experimental chemist. . . but she realized we really needed to have a computational program here if we were going to stay competitive. She wanted to make sure we’re really pushing the envelope of what could happen at the high school level.”

King, who had no exposure to computational science until her junior year of college, still finds it amazing that she is teaching computational science in a public high school. “It’s not [even] a required part of the curriculum at the undergraduate or graduate level,” she says. “You could go through and have your Ph.D. in chemistry and never have been exposed to concepts in computational chemistry.”

Science and Math students are working with information from some very heavy hitters. “I’m always looking for new data sets to analyze” in the classroom, says Payne, who pulls data daily from CERN, the world-renowned European research organization that houses the planet’s largest particle physics laboratory. He also draws data from LIGO, the Laser Interferometer Gravitational-Wave Observatory run by Caltech and MIT.

“I didn’t realize. . . how much was possible,” Em says of her computational courses. “The models are getting better every day with more people putting data into them . . . so we are able to really effectively model these systems.” 

Stepan Malkov, from Ardrey Kell High School in Charlotte, is a senior in NCSSM’s residential program. Before coming to Science and Math, Stepan was interested in pure mathematics — the study of the basic concepts and structures that form the foundation of mathematics. “When I came here I heard a lot of good things about the computational program,” he says, “and I realized it does relate a lot to mathematics because the fundamental structure behind the processes and algorithms is inherently mathematical.

“I found it really interesting,” he continues, “how I might look into those algorithms and kind of try to find patterns within them so that I can optimize them and add my own perspective to those techniques.” 

Stepan enrolled in Gotwals’ Research in Computational Science course, an undertaking that bridged his junior and senior years and kept him busy during the summer in between. The course taught him how to apply computational research methods to research projects in nearly any discipline. As with Em, the introduction to the field expanded Stepan’s vision for the future. 

“Right now I’ve shifted my interests more from pure mathematics to applied computational mathematics,” he says. “I definitely realized that that is a field I was not aware of before that has very interesting applications in any mathematical or modeling type of problem. It doesn’t really matter what your field is; what matters is you can apply those computational techniques in that field.”

Such as economics. Stepan used computational research skills to investigate the role different macroeconomic variables, like employment or inflation rates, played in fluctuations within the stock market. He was surprised by the results. “I didn’t think I would get something that would agree with what reputable Ph.D.s and other scientists did, but it turned out that not only did my results agree, but they actually expanded on them and they offered, kind of, new ways of interpreting those results that can be used to gather the significance of what emerging variables are important in computational econometrics. 

“It’s an eye-opening experience,” Stepan says of the NCSSM introduction to computational science. “It shows you the opportunities that you have at this school that you might not have at any other school.”

In the humanities, text as data

The power of computational science isn’t confined to traditional scientific fields like chemistry, biology, or physics. At NCSSM, it has crossed into the humanities as well. Gotwals and Michelle Brenner, an instructor of humanities at NCSSM, co-teach a course in the residential program called Digital Humanities.

“Digital Humanities is sort of this big umbrella term that we’re using for a lot of different things, both here and the collegiate level right now,” says Brenner. It includes, she explains, not only how humanists digitally archive the things they may need for their research — census data or the narratives of enslaved people, for example — but also the computational methods they use to examine that data for meaning. At NCSSM, that means using “text as data.” 

“When we start pulling apart text as data, it goes from the sort of close reading that we do in our more traditional humanities classes, to what they call ‘far’ reading, or ‘distant’ reading,” Brenner explains. The advantage of far reading, she says, is that it allows scholars to look beyond the word on the page to see how it relates to other cultural or societal elements. She points to a recent project by one of her Digital Humanities students that explored Google search trends, Wikipedia posts, and Reddit threads to analyze the difference in perception of mental health between the platforms, particularly with regard to national news events such as mass shootings or presidential elections. 

Students look to texts within the traditional literary canon as well; Brenner cites the computational exploration of Charlotte Bronte’s revolutionary Victorian-era novel Jane Eyre. “We’re not looking at Jane Eyre as it stands by itself,” Brenner says, “but we’re looking at the words in Jane Eyre and how they compare to all of the other novels written in the same year.” Such comparisons can shed additional insight into societal markers such as gender, race, or social status common to a particular period. 

Brenner and other literary scholars steeped in traditional approaches were, as might be expected of those drawn to the weight of a book in hand, skeptical of computational science’s usefulness in humanities. They feared that the application of data science would overlook important, intuitive understandings available only through close reading of a text. “I thought that close reading was the only way to figure out what a text has to say,” she admits.

Now, “it’s just another tool that humanities scholars can use,” she says. “It lets us take the same texts that we are familiar with and look at them in a different way.”

Strangely enough, computational science’s entry into the humanities has had the ironic effect of reinforcing the ambiguity inherent in the study of a society and its culture. “I think, more often, digital humanities complicates the claims we’ve always assumed,” Brenner says, “which is kind of what humanities likes to do.”

A tool for new breakthroughs

Computational science is neither flawless nor absolute. It won’t displace existing disciplines, nor will it be rendered ineffectual by them. But it can help scientists of all stripes get closer to a commonly desired outcome, be it a film on a microchip that revolutionizes communications, a new drug to treat cancer, or a solution to threatening environmental issues. “We will make it farther in our [respective] industry,” King says, “if we all support each other and build off the progress of each other.”

And while this “new” kind of science continues to find its way into disciplines beyond the hard sciences, its greatest achievements will likely be realized in more traditional fields of investigative research where many of the grandest of Grand Challenges reside. 

“Most of the Grand Challenge problems are computational in nature,” says Gotwals. “They’re probably not going to be solved using experimental methods [alone]. I’m not bad-mouthing experimental scientists. . . but the breakthroughs are probably going to come from computational” discoveries that are then verified and realized in the lab.

If you can solve even a small piece of such puzzles, says Bob Gotwals, “you go to Sweden and pick up your Nobel Prize. The new frontier is computational. We want our kids prepared for that, we want them ready for that.”