ÃÛÌÒÊÓƵ

Researchers like Kalivas, who devote their life’s work to the niche field of chemometrics, use mathematics, statistics, computer science, and techniques from other fields to learn all they can about chemistry through data.  

“It scratches my creativity itch,” said Kalivas. “I can create algorithms that can mine through data and find the chemical relationships that other people can't find and help solve big, real-world problems.” 

Kalivas was at the Fourth International Conference on Environmetrics and Chemometrics in September 2000. Held in Las Vegas, he had been invited to share his take on graphical data analysis – think charts, graphs, and the like – in a session called "Multivariate Data Analysis and Visualization." About 50 people gathered in a conference room to hear Kalivas and four other experts share their insights and take a few questions from the audience. One question lasered in on what the panel thought the future held for data visualization. It was then that Kalivas hopped up on his soapbox and issued his proclamation, vexing some of his colleagues.  

“I said, ‘I see that a big paradigm change would be to do your data analysis in virtual reality,’” stated Kalivas. “A chuckle arose, and I didn’t know how to take that.”  

The idea had been needling at him for some time before the meeting. Kalivas had seen the dystopian sci-fi film Johnny Mnemonic, and in a scene, the data courier Johnny, played by a young Keanu Reeves, uses a virtual reality system to sift through data on what folks today would call the cloud. While the futuristic VR hardware used by Johnny looks dated by today’s standards, at the time, it provided Kalivas with plenty of inspiration. 

“Johnny Mnemonic gave me the idea, and I said, ‘I want to do that. I can do that. We can do that,’” Kalivas explained. “Ever since then, it’s been a career ambition.” 

Kalivas wasn’t just a young-buck researcher who happened to have a light-bulb moment or an old crank yelling at the clouds when he pitched his sci-fi-inspired idea. He was near the midpoint of his career and had established himself as a leading mind in the discipline. Kalivas earned his doctorate in analytical chemistry in 1982 from the University of Washington under Bruce Kowalski, one of the founding fathers of chemometrics, who pioneered the field in the early 1970s. 

”His level of excitement was always at 10-plus, and it was very catchy and made you want to go. His ability as a motivator was incredible, and his passion wore off on me,” Kalivas said of his mentor.    

For his dissertation, Kalivas worked on a three-pronged project with Kowalski. First, he was tasked with developing a multivariate calibration method. Multivariate calibration methods allow analytical chemists to use multiple data points to predict the value of another point. However, these methods are only as good as their ability to correct for what researchers call matrix effects. Matrix effects are the combined impact of all the different parts and pieces in a sample that can cause changes in the specific item a researcher wants to measure. For the second part of the project, he created an algorithm that accounted for the multiple matrix effects in the data simultaneously, leading to more accurate results. Lastly, he used assembly language – binary computer code that uses just 1s and 0s – to connect a computer and a spectrometer to do automated analysis of samples.  

“That was hardcore computer science,” said Kalivas. “For my Ph.D., I had to pass a foreign language proficiency exam, and I argued in front of the committee that I should be able to use that assembly language as my foreign language proficiency. They didn’t go for it, and it had to be a true foreign language.”     

After his time in Seattle, Kalivas moved to the Midwest for an instructor gig at the University of Minnesota Morris, where he taught several courses, including general chemistry, instrumental analysis, and computers in society. He then went south to Texas A&M University, where he lectured in analytical chemistry. In 1985, he landed a tenure-track faculty position at ÃÛÌÒÊÓƵ and established his research lab, focusing on analytical chemistry and chemometrics. From his home base in Pocatello, the early research out of the lab focused on multivariate calibration and how to obtain the best information from datasets using computers, tackling problems ranging from detecting impurities in pharmaceuticals and determining if a biopsy is cancerous or not, to how ripe a fruit is and the age of a fish for sale at the market, and many more.

“I like to get the idea and visualize the molecular dynamics first,” Kalivas said. “Then talk about it with my students. Then we get to the math and try to model it. Once you have the mathematical description of what we’ve verbally expressed, we code it and test it on our different datasets to see what we're missing. To diagnose our problems with the code, we create a lot of graphical output, and we can see what’s going on in the code.”

And because chemometricians borrow from multiple specialties to complete their work, it’s reflected in the students who work in Kalivas’ lab. Besides the chemistry majors you’d expect, his students form a patchwork of majors from across the STEM fields. Like a college football coach scouting for players at high school games, Kalivas would use the general chemistry courses he taught to spot the students who could be a good fit for working in the niche laboratory. 

“I want students who want to be researchers and not a pair of hands,” said Kalivas. “They need to be motivated, eager to learn, collaborative, and creative.”

Once he’d found his recruits, Kalivas would offer them a chance to meet, learn more about the work involved, and possibly offer them a job. One of those students — and one who was on a very different career trajectory at the time — was a physics major named Robbie Spiers. 

“I thought that I wanted to study physics in undergrad and then become a patent lawyer,” said Spiers. “During that very first meeting, where I didn't even know what undergraduate research was, Dr. Kalivas talked about his many different research projects: rapid quantification of moisture and protein content in corn samples, classifying different types of microplastics that can end up in the environment, restoring defaced serial numbers on weapons, authentication of Italian olive oil, and much more.”

Legal aspirations aside, Spiers took him up on his offer and dove into the work, and excelled. Over the next three and a half years, he published six papers, gave four talks, presented nine posters around the country, won three awards for his research, and was named a winner of the Outstanding Student Achievement Award by ISU’s College of Science and Engineering. Kalivas gave Spiers his highest of praises, saying in 2020 that Spiers was “the best I have seen work in my lab in the last 35 years.” Spiers, who has since started his doctorate at the University of Delaware studying , is quick to point back to Kalivas for his success.  

“Mentoring undergraduates to the level where they can contribute state-of-the-art research in a rapidly advancing field is an incredibly difficult task, but one that Dr. Kalivas has an incredible ability for,” said Spiers. “He gets new students up to speed on the research, starts them out on interesting projects, and continues to nurture with weekly meetings, where he makes sure that students have a strong understanding of the material and an ability to communicate the science to their peers.”

After his early work on multivariate calibration, Kalivas shifted the lab’s focus to specifically matrix effects, those pesky things that can mess with the measurements of the item being studied by a researcher. Soil samples have some of the most challenging matrix effects to overcome because of how wildly different soil can be, inch by inch. 

“For example, if you dig up some dirt where your dog goes to the bathroom, that’s totally different from the dirt two feet over or the south side of the yard versus the north side of the yard. It’s totally different soil.”

Matrix effects are the nemesis of analytical chemists, says Kalivas, and his shift in focus was driven by a change in his thought process on them. 

“I realized it’s not so much that we want to get rid of (matrix effects) as we need to leverage it and have it work for us, said Kalivas.   

As he has progressed throughout his career, each project and each pivot – from his work on multivariate calibration to matrix effects – has been an incremental step toward his Johnny Mnemonic-inspired vision. For the past 15 years, his research has focused on model updating.

“Say I have a model that can predict the amount of protein in a certain brand of milk,” explained Kalivas, “Now, I want to use that model on another brand of milk, but it doesn’t work because it’s a different brand, which has a different composition. Getting a model to work on the new brand is called model updating.” 

His work over the years on model updating has led to the creation of models that can determine the purity of olive oil, identify the elements in the soil, and detect the active ingredients in pharmaceuticals.

“In the process of updating a model, you create hundreds and thousands of new models, and then you have to figure out which one is the right one,” Kalivas continued. “By default, I said, ‘We have to figure out an algorithm to find the right model,' and we did.” 

Algorithms, however, are only so good at pattern recognition, and since they have rigid parameters, they often miss outliers on the edges. It was this work that led to a project Kalivas thought could be tackled in virtual reality.

“Humans have been trained since childhood to recognize patterns, spotting the difference between seemingly the same pictures is a good example,” explained Kalivas. “I thought this could be our first VR project: we put these hundreds and thousands of models in VR and let the human figure out what the right models are.”

This brings Kalivas now to his pièce de résistance in terms of his research career. In September 2023, the National Science Foundation funded his project, “Immersive Virtual Reality for Discovering Hidden Chemical Information and Improving Multivariate Modeling and Prediction,” to the tune of nearly $450,000 over three years and helping bring his dream of chemical analysis in VR to life. 

"I was obviously excited to receive the funds to carry out my capstone career project,” said Kalivas.  “This will be a project that can potentially open up many new research directions in chemometrics as well as other fields."

The following month, Kalivas debuted the first movement in his magnum opus: Uncovering Chemical Information Using Virtual Reality. The talk, co-authored with his student Jordan Peper, saw Kalivas return to Nevada, the state where he made his bold prediction more than two decades prior, and showcase his efforts on how VR can help people make quicker and more accurate model selection decisions. 

"I was nervous presenting the talk,” Kalivas said. “I think as we publish more papers about the applications, the utility of VR will become more apparent.”

2024 saw Kalivas retire from the teaching side of being a professor and focus solely on his work in chemometrics in VR. Since then, his talk with Peper has been accepted in paper form by the scientific journal Applied Spectroscopy and is scheduled for publication in 2025, along with five other papers, some of which are already published or in the review process. In October 2024, he presented a functional VR tool for outlier detection at the SciX conference in Raleigh, North Carolina. In the VR system, the data is presented in an abstract form called glyphs. Resembling a medieval mace, these glyphs are spheres with varying sizes of spikes on them, and the bigger the spike, the more of whatever a researcher is measuring has been measured in their samples. Rather than pore over a spreadsheet or a graph, a researcher using the VR system can quickly, accurately, and easily see which samples are outliers amongst all their data. 

"These talks and papers only scratch the surface of the power of VR in chemometrics and data analysis in general,” said Kalivas.

This summer and thanks to additional funding by the National Science Foundation, Kalivas and a graduate student will be adding haptics-think of the vibration you feel when your finger touches the screen of your phone–and sonification–turning the data in to sound– to the VR glyphs, meaning researchers who are blind or visually impaired, or deaf and hard of hearing, can utilize the system.

"The blind students we have interviewed and agreed to work with us are ecstatic about actually being able to do data analysis,” Kalivas said. “Creating tools allowing blind people to expand their knowledge base provides a genuine feeling of serving society, something that there is always a greater need for."

He’s also further cementing his status as a trailblazer in the field by editing a special issue of the Journal of Chemometrics titled Immersive Analytics with Virtual Reality: The Frontier is Here.

“Because VR is new to chemists for data analysis, most of the contributing authors are computer scientists," said Kalivas.

All told, and when he wraps up the project in 2026, Kalivas’ hopes are simple: inspire the next generation of chemometricians to take things a step or two further. 

“VR technology is readily available,” Kalivas said. “By the end of the grant, we hope to have cracked the VR nut, allowing young scientists to develop other VR paths to advance data analysis.”

The coming years of work in his lab will see Kalivas tie a fitting bow on his research endeavors and his bold prediction for the future come to fruition. The VR project will see Kalivas working with what will be his last student. Over the years, he has mentored nearly 70 undergraduate and graduate students in his lab and taught scores more through his classes. Beyond being a bit of a visionary, it’s the students and minds he’s left a mark on that will be Kalivas’ legacy at ÃÛÌÒÊÓƵ State and beyond. 

“I have a million stories I look fondly back on: from the backyard barbeques with the lab, a long road-trip to Palm Springs for a conference in 2019, skiing together at Grand Targhee, some three-hour long one-on-one meetings where we'd look at hundreds of research figures and write all over the whiteboards to discuss our theories, and so, so, so, so, much more,” said his former student Robbie Spiers. “Working with John was maybe the most singularly influential experience in my life, affecting my career choices and prospects in addition to my overall outlook on the world. I can't thank him enough.”

For more information on ISU’s Department of Chemistry, visit .