Kurt Vonnegut once called science “magic that works.” Considering the implications of the work that Bronx Science research students have conducted, science truly has taken on a miraculous lens, turning hundreds of hours of dedicated study into an almost limitless future in medicine, media, and more.
Many students demonstrated this power through participation in the Regeneron Science Talent Search, or “Regeneron” for short. Awarding $1.2 million to 300 students across 36 states and China, Regeneron serves as the oldest and most distinguished science and mathematics competition for high school seniors.
Regeneron highlights studies with worldwide impact, whether through models working towards a cure of celiac disease to robots and satellites to governmental analysis. Past honorees have since received Nobel Prizes, National Medals of Science, and similarly prestigious awards.
Out of more than two thousand applicants across 712 schools, Bronx Science has nine “scholars” — the term for semi-finalists — and the second highest number of selected students of any school featured. They have been recognized for their impressive research, innovation, leadership, academic rigor, and community engagement.
Let’s honor this year’s Regeneron scholars: Sophie D’Halleweyn ’24, Aiden Hightower ’24, Melody Jiang ’24, Nema Khan ’24, Ryan Kim ’24 (Official Class D20), Jonathan Lin ’24, Dimitrios Mahairas ’24, Kun-Hyung Roh ’24, and Rachel Wu ’24.
D’Halleweyn and Roh were also named finalists, now competing for further awards in a weeklong extensive judging process with a guarantee of at least $25,000 each. Bronx Science is one of three schools nationwide with the honor of having more than one finalist, a truly prestigious recognition.
Since the inception of the program eight decades ago, Bronx Science students have undergone a three-year path of intensive research projects, culminating in Regeneron submission. They collaborate with professors, researchers, and doctors — often more than double their ages — to produce tangible and meaningful studies whose impact spans far beyond the classroom.
Entering research at Bronx Science is a lengthy and rigorous endeavor, beginning with the Science Research course and outstanding grades in ninth grade to qualify for this individualized program. There are four sections: social sciences, mathematics and computer science, physical science/engineering, and biology. Students pick a track to begin in sophomore year, continue developing their projects as juniors, and then submit final papers in their senior fall.
The past four years have not reflected the typical process, however. Dr. Tracy LaGrassa, Biology Research teacher, shared, “Most students in this group of seniors experienced their first year of high school remotely, during the height of the pandemic. They did not have the benefit of 9th grade research — the foundational course of the research program — and it was really hard to get a lab after sophomore year, because of COVID restrictions on lab personnel. Yet, they still managed to make all this progress. They came to school in 10th grade, formed real friendships, and got deeply involved in the school culture, research as only one part. They helped their communities. They had tremendous responsibilities at home. How they did all this, I cannot imagine. I am beyond proud of them.”
Dr. Vladimir Shapovalov, Mathematics and Computer Science Research teacher, reaffirmed this statement, identifying the scholars as exhibiting “diligence, intelligence, and integrity.”
Regardless of the many obstacles, Bronx Science research students — the scholars in particular — have remained steadfast in their commitment to their work, developing skills from creating crystals to finally breaking down the mysterious formatting of academic journals. Now, let’s explore each of their journeys more individually:
Sophie D’Halleweyn — Alleviating the Energy Crisis: A Novel Multi-Task Machine Learning Algorithm for Designing Efficient Nanocatalysts To Reduce Industrial Energy Impact
D’Halleweyn designed a machine-learning algorithm — multi-tasking algorithm for variational encoding, or MAVEN — using XAFS (a type of X-ray imaging) to characterize catalysts. Named after the Yiddish word for expert, this program created a vector space to reflect the physicochemical properties of palladium and palladium hydride atoms. The algorithm can determine the most efficient nanocatalyst to optimize reactions.
She shared, “My project addresses a lack of explainability and accuracy in previous machine learning algorithms which analyzed nanocatalysts. In addition, it overcomes the lack of information in the near-edge region of X-ray absorption fine structure spectra (XANES); the extended region is defined by Fermi’s golden rule, but before MAVEN, there was no technique which could conclusively find usable information from XANES.” This can be applied to almost all electronics, hydrogen, batteries, and even energy sources as a means of combating climate change.
Aiden Hightower — Molecular Fusion: A New Methodology for the Rational Design of Novel DNA Motifs via the DX-Mediated Binding of DNA Tensegrity Triangle Unit Blocks
Hightower has combated limitations of designing DNA crystals by finding a new methodology for novel motifs from DNA using the tensegrity triangle, the “field’s favorite unit cell.” In the process, he created his own crystals.
Hightower stated, “Though my work is very structural, it provides an eventual means for future scientists (even those outside the field of nanotechnology) to design crystallographic motifs with ease. Nanocrystals are useful for many reasons, the main being suspending and protecting more important chemicals within. That is, for quantum computing, nanoelectronics, drug delivery, cell-protecting, and pretty much any ‘small’ application you can think of. Overall, it makes DNA crystal design a lot easier and a lot more expansive.”
Melody Jiang — Characterizing Effects of Natural Killer Cell Neural Cell Adhesion Molecule 1 Interactions With Stromal Cells on Natural Killer Cell Development
Jiang studied how a certain immune cell — Natural Killer Cell (NKCs) — interacts with bone marrow-derived mesenchymal stromal cells (MSCs), thereby encouraging the former’s development.
Jiang explained, “Natural Killer Cells make up about 10-20% of our body’s circulating lymphocytes so although they are very important, not much is known about them. Primarily, Natural Killer Cell development remains elusive — especially the interactions that support its development. My project aimed to decode one of such interactions. Originally, we knew that NCAM1 was the protein that was on NK Cells that were interacting with MSCs. What we did not know was what protein on MSCs the NCAM1 was interacting with. I hypothesized that it was also NCAM1 on MSCs, as NCAM1-NCAM1 interactions between NK Cells and other cells had been previously observed.”
NKCs are responsible for bolstering the immune system and understanding their cell development effectively — as Jiang has done — may prove crucial for cancer treatment and the attacking of tumor cells.
Nema Khan — The Impact of Media Attention to Gun Violence on the Elevation of Mass Shootings: The Rise of Copycat Shooters
Khan analyzed the effects of news reports on occurrence of mass shootings, concluding that a rise in news reports with the words “mass shootings” may precipitate an increase in mass shootings.
She shared, “Through my research, I’ve discovered that the copycat phenomenon exists in relation to mass shootings, presenting a new perspective to previous studies which primarily focused on how social media attention augmented serial murders.”
Motivated by the prominence of mass shootings in the U.S., she hopes her work will encourage media and news outlets to enact more careful guidelines when relaying information about mass shootings, reducing the influence on potential perpetrators.
Ryan Kim — Exploring the Pathways of c-Maf Endothelial Cell Reprogramming
Kim studied how combinations of cytokines and signaling molecules can enhance the reprogramming of endothelial cells, potentially altering how liver diseases and injuries are treated.
He summarized, “My project aimed to improve the efficiency of reprogramming generic vascular cells to differentiate into a more specified version that is crucial for liver regeneration. This is crucial for liver regeneration therapies. I used a method called lentiviral reprogramming, introducing a transcription factor to encourage their transformation.”
Jonathan Lin — Analysis of Distinctive MicroRNA Conservation Patterns as Markers for Unique MicroRNA Processing Mechanisms
Lin investigated evolutionary patterns of microRNAs for 10 closely related species in the Euarchontoglires superorder of mammals, discovering five distinct classes. The improper regulation of microRNAs is connected to cancers and malignancies, so studying their characteristics — and the differences in gene regulatory networks across species — better informs the understanding of related functions.
He shared, “This highlights that we still have much to learn about how microRNAs are regulated within all organisms, but most importantly, humans. It shows that there are a lot of hidden actors in the microRNA regulatory network that we do not yet understand. I hope that my discoveries can lead to future investigation about how these regulatory mechanisms can be manipulated as cancer therapies.”
Dimitrios Mahairas — Overcoming Dataset Imbalance Using Light-Weight Vision Transformers With Applications to Computational Drug Design Derived From Botanical Prototypes During Post-Antibiotic Era
Mahairas designed a comprehensive machine learning model to identify millions of plant species. By discovering, categorizing, and understanding the structures of images, researchers can investigate plants’ medicinal properties, develop new antibiotics, and save millions of lives in the process. The algorithms can also be used to identify invasive plant species, therefore protecting ecosystems and providing stability to infrastructure that might be otherwise threatened.
Mahairas explained, “Machine learning algorithms are a way to make a computer think like a human. For example, you give the computer an image of a cat, and the computer will tell you that the image is of a cat. In my research, I dealt with algorithms which were specifically used to identify the species of a plant in an image. Just like a human with an expertise in taxonomy, the computer could take an image of a laid out plant and tell you the species of that plant. The difference is, a computer can do this with thousands of images at a time…In my research project, I aimed to create a model architecture that performs accurately with these plant image datasets, while also being lightweight and accessible to all researchers, regardless of their processing power.”
Kun-Hyung Roh — Novel Drug Discovery Methodology Using Machine Learning for Gene Expression-Based Virtual Screening Predicts Novel Compounds To Reverse Alzheimer’s Disease With Applications to Cancer and Longevity by Inhibiting CtBP2 Expression
Roh built on his lab’s studies on phenothiazines, a type of drug counteracting the inflammation that drives Alzheimer’s Disease (AD), by incorporating a machine learning model to demonstrate the progression of the drugs and study the gene expression of CtBP2. He discovered the amplified presence of the protein with older individuals, especially in the brains of those with AD. Furthermore, he encouraged a paradigm shift for drug discovery techniques and used AI to predict drugs that reduce CtBP2 and even outperform phenothiazines. His studies reflect mechanistic and neuroprotective properties of phenothiazines not previously understood. The ramifications of his work extend to improving the accessibility and efficiency of treating age-related conditions, namely AD and cancer.
Roh shared, “My grandparents took care of me during my childhood. I lost my grandfather to Alzheimer’s Disease, and my grandmother is losing memory to Parkinson’s Disease. These diseases unfortunately do not have a cure nor satisfactory treatments. In honor of my grandparents, I wanted to find a cure for such neurodegenerative diseases so that others would not have to go through the same pain that my family experienced. I hope that in the future I could develop a cure against age-related diseases for everyone.”
Rachel Wu — Detecting the Effect of Textual Features in Social Media Using an Innovative Machine Learning Approach With Applications for Managing Public Opinion
Wu studied the ability of content on social media to influence its audience, analyzing the paths of misinformation, bolstering the impact of socio-political and similarly time sensitive issues, and identifying features that can detect fake news. She created a standard with sentiment analysis that can be applied in marketing, political discourse, and more. She also distinguished the length of text in a post — first on Twitter and then Reddit — as the feature most important in making them influential to readers.
Wu reflected, “I noticed that some texts have the capacity to reach and resonate with a wide audience while some fall short in their impact, which is crucial for global issues. Therefore, I conducted this research project to identify the distinctive features that make certain texts remarkably influential. By understanding the features that enhance the effectiveness of certain texts, more robust strategies can be identified to maintain the integrity of the online global community.”
Emerging from the Lab
All scholars had very different approaches to entering research. Melody Jiang shared, “I have always loved biology and knew that I wanted to become a researcher since my very first science class in the third grade. I came to Bronx Science to pursue science and join the research program because I knew it was the next step on my path of pursuing research as a career.”
In contrast, Dimitrios Mahairas joined research as a junior, having heard about the program from his sophomore math teacher. Regardless, the underlying drive to answer difficult questions and improve the world, unbeholden to the timeline or shape it takes, remains a constant for these students.
The scholars have adapted over the course of their studies, developing skills from the typical data analysis to the less technical but still crucial successful subway navigation, and built projects that will radically change the future of science. They have shown intense dedication, exemplified by their willingness to spend hundreds of hours in lab for the sake of research. Their age is an inspiration, equipping them to approach research with new ideas and creative problem-solving.
As Kun-Hyong Roh shared, “I was so excited when I finally got the chance to join the research program where I could bring my own thoughts on experiments into reality. The primary lesson was that research is a continuous process of learning and studying. It is an endless test of patience that demands motivation and passion. To pass these tests, I had learned to search for ‘why?’ behind anything I do. Why do I conduct research, why shall I pursue this study, or why do some activity? I learned that breakthroughs and achievements can come secondary. I realized that what’s more important to me is this intrinsic ‘why,’ and by unveiling the ‘why,’ we can reach discoveries and innovation that go beyond earning medals and trophies.”
Furthermore, from running to reading Russian literature, all nine scholars engage in a diverse range of hobbies that inform not only their interests but their studies. Sophie D’Halleweyn shared, “I’ve loved dancing for as long as I can remember. I’m classically trained in the Vaganova method of ballet, which has taught me [so] much of the discipline and dedication I needed to succeed in research.” Research is multi-dimensional, rooted in passion and commitment, which the scholars hold in spades.
All intend to continue immersing themselves in scientific pursuits after graduation, and one can only imagine excitedly what next steps they will take. Dr. Scott Savaiano, Social Science Research teacher and mentor, encourages Bronx Science students to follow suit: “Don’t hesitate, do it! You deserve the opportunity.”
As Kun-Hyong Roh shared, “I was so excited when I finally got the chance to join the research program where I could bring my own thoughts on experiments into reality. The primary lesson was that research is a continuous process of learning and studying. It is an endless test of patience that demands motivation and passion. To pass these tests, I had learned to search for ‘why?’ behind anything I do. Why do I conduct research, why shall I pursue this study, or why do some activity? I learned that breakthroughs and achievements can come secondary. I realized that what’s more important to me is this intrinsic ‘why,’ and by unveiling the ‘why,’ we can reach discoveries and innovation that go beyond earning medals and trophies.”