When I first began my IDEA Grant research project, I knew I was ultimately undertaking a project that sits at the intersection between biology, computer science, and mathematics. My project, which focuses on creating a new computational tool for analyzing spatial transcriptomics data, is not just solely about writing code or producing pretty figures. At its heart, it is about addressing one of the fundamental challenges of modern biology: how to capture and understand the organization of life at the multicellular level.
The impact of my work lies in how it contributes to the rapidly growing field of spatial transcriptomics. Traditional single-cell RNA sequencing provides us with a powerful view of what genes are expressed in a cell, but it removes those cells from the tissue context they came from. To create a similar example in real life, this would be like trying to understand a city by listing the characteristics of its inhabitants without actually knowing where they live. Spatial transcriptomics brings that missing dimension back, letting us see where cells are in tissue and how they interact. My project, the creation of a method called Diffusion Spatial Pseudotime (DSPT), is an effort to improve how we trace the “trajectories” of cells, as in, how they transition from one state to another while accounting for their spatial positions.
This puts my work in dialogue with leading computational methods such as stLearn, which combines spatial and morphological information to reconstruct trajectories in tissue, and SpaceFlow, which integrates expression and spatial data to build spatiotemporal maps of development and disease. Where those tools use machine learning and deep graph networks, my approach is more mathematically rooted in diffusion processes. Together, all of these approaches form a conversation in the field: how can we best capture not just the static organization of tissue, but the dynamic, evolving relationships among cells? My work adds a new voice to that conversation, offering a different angle on how diffusion-based methods can reveal cell organization patterns.
If successful, this research could have effects beyond computational biology. Improved methods for analyzing spatial data can help cancer researchers understand how tumors grow and interact with immune cells, or help neuroscientists trace developmental pathways in the brain. It’s not hard to imagine how a more precise tool for mapping cell transitions could help identify drug targets, design therapies, or simply accelerate our ability to generate biological insight from massive datasets. Even for other computational researchers, my work offers another model to compare against existing benchmarks, a continuous cycle of pushing everyone forward by clarifying what works, what doesn’t, and why.
As I return to campus at FSU, my immediate goal is to finalize a polished version of my tool: organizing the code, documenting it clearly, and making it publicly available so that others can use and test it. I also plan to prepare my results for presentation, compiling the figures and metrics that demonstrate the method’s strengths and limitations. This will serve not only as the culmination of my IDEA Grant but also as a foundation for my undergraduate honors thesis.
Looking further ahead, I am excited to build on this experience as I prepare for graduate school. This project has shown me how much I enjoy the process of asking technical, open-ended questions and then engineering solutions to answer them. Beyond FSU, I hope to continue in computer science research with a focus on computational modeling within biological contexts. I plan to apply to graduate school this upcoming admissions cycle, and I think this project is a great demonstration of the abilities I have gained over the summer, that is, being able to conduct effective and focused research.
More than anything, this research experience has given me confidence. I’ve learned how to navigate scientific literature, how to benchmark my work against established methods, and how to stay persistent even when the code doesn’t run the first ten times. I’ve also learned how to find my own academic “voice”, how to argue for why my method matters and where it fits in the bigger picture. Whether or not DSPT becomes widely adopted, I know that I’ve made a contribution that others can build on. That, to me, is the essence of research: adding one more piece to the puzzle so that collectively we can see a clearer picture of life’s complexity.
HSF House Blogs 