On January 8, 2025, Benjamin Gitai ’25, Amy Lin ’25, and Yurai Morales ’25 were named Regeneron Science Talent Search Scholars, each winning $2,000, with an additional $2,000 going to PHS. Morales was further selected as a top 40 finalist, entitling her to an award of at least $10,000. The Regeneron Science Talent Search, one of the largest research programs for high school students, selects 300 scholars annually among 2,500 applicants, with the center of the application being a 20-page research paper.

Benjamin Gitai

Benjamin Gitai’s ’25 research focuses on improving the success rate of ankle replacement surgeries.

“We can [build] a model to essentially ... try to predict if [failure] will occur and then be able to help stop that from happening,” said Gitai.

Gitai was inspired to begin the project after attending Princeton University’s Laboratory Summer Learning Program, a free program for high school research students. At Princeton, Gitai was part of a group that focused on ankle imaging led by professors Dmitri Kosenkov and Olga Troyanskaya. He saw that the orthopedics, doctors dealing with deformities of bones and muscles, measured the ankle by hand.

“[An] orthopedic surgeon [was] telling us about how they take all these measurements, and it just struck me that with all this technology that they still have to [measure by using] these little points [on] this X-ray machine. I thought I could solve that issue and help them with that. From there, I was able to expand that [idea] into the prediction algorithm,” said Gitai.

Because the idea of using AI to measure ankles is relatively unexplored, Gitai started off his project by studying on websites like Stack Overflow, a question-and-answer site for programmers, and Github, a platform for storing code.

“I did lots of research into what [code] libraries and various tools I should be using. I’ve [had] a lot of experience coding, but I didn’t have experience using any of this computer vision material. I had to learn and teach myself all of that from scratch,” said Gitai.

However, despite his success at Regeneron, Gitai still considers himself at the beginning of a longer journey. Gitai is planning on attending Princeton University next year, and hopes to further his research by using the university’s resources.

“I think that’s one thing that makes this project a little bit special is the fact that it has so much room to grow. We’ve created the bone detection model, and right now, we’re waiting on getting data to be able to further expand the failure prediction. But the truth is that the pipeline that I created applies for literally any bone in any kind of bone deformation. Potentially, this [project] could be expanded indefinitely,” said Gitai.

Amy Lin

Amy Lin ’25 didn’t expect to be named a Regeneron scholar.

“It was a very pleasant surprise because I went in there thinking, ‘oh there are some pretty cool projects that they always select, year after year.”

But in early January, Lin was notified that her project, titled “Predicting Compound Melting Temperatures From Computationally Derived Properties Using Machine Learning,” had been selected. Her research focused on using AI to predict the melting temperatures of different materials by analyzing their material properties.

“The problem I was trying to tackle was that, right now, there are a lot of methods [for] experimentally measuring melting temperature [of] different chemical compounds. And those [methods] are pretty dangerous and pretty costly. And existing simulations for computing melting [points] are also pretty time-consuming and computationally expensive,” said Lin.

Inspiration for this project came from The Coding School’s data science summer research program.

“I learned about this topic, which is predicting melting temperatures. I wanted to actually continue this project because I fell in love with machine learning and data science. I wanted to make some contributions to the area and perhaps develop a toolkit to predict melting temperatures and other such material properties,” said Lin.

After a bit of trial and error, Lin settled on the Random Forest Regression prediction model, which predicts melting points by anaylizing other properties with the high correlation to melting point.

Throughout the process, Lin relied on PHS research teacher Mark Eastburn and science supervisor Jacqueline Katz.

“They really created a great environment for [me] to explore [my] project. They offered a lot of guidance in terms of what [are] good research practices? Or, how do you annotate and read papers properly? Or, how [can we] write our own reports? So that was really helpful,” said Lin. Lin emphasized that the Regeneron competition was an honor that has helped develop her appreciation for AI and computer science.

“I think the project definitely bolstered my interest in these areas [computer science], because it showed me the power of what machine learning and AI can do to help us learn about the world around us. It’s really an amazing tool,” said Lin.

Yurai Morales

Yurai Morales ’25 has had an unconventional journey to the highest level of high school research. Morales was born in the United States, then moved to Guatemala at age seven, then to Mexico at age 14, and finally back to the United States at 17, when she enrolled at PHS. At PHS, research teacher Mark Eastburn encouraged Morales to develop the work she began in Mexico into a research article.

Her final project, titled “How the Lion Becomes a Lamb: Transfer of Bacterial Symbionts From Ant Larvae to Vegetarian Spiders Through Selective Predation​​​​​​,” was announced as one of the top 40 finalists on January 23th, the highest award a PHS student has received in the competition. As a finalist, Morales will be going to Washington, D.C. on March 11 to showcase her research to the United States’ leading scientists for the chance to win a top 10 prize.

In her research, Morales discovered that the species Bagheera kiplingi, which is the only spider known to be a herbivore, is able to maintain its plant-based diet because it eats the larvae of ants. Morales also investigated the bacteria that allows the spiders to extract nutrients from the larvae.

“I was wondering why these curious spiders could be vegetarian when their bodies were designed to be carnivores. I was doing this research to understand how their body works and how they can eat plants without having a carnivorous life,” said Morales.

One of the biggest challenges for Morales throughout the project was not just the research component , but learning English.

“English is not my first language. Some of the articles I read [are] in Spanish, but I have to translate [into] English, so it’s very hard for me. I’m still learning,” said Morales.

Morales hopes to continue her project at The College of New Jersey, where she is enfrolled for this fall. But for now, Morales is proud of the work she has already accomplished.

“I’m feeling very excited. It was very surprising for me; It’s such an amazing opportunity for me because I’m learning new skills and doing research. English is my second language, but I’m still here, and I’m very proud,” said Morales.