Introducing a new Research Internship Student of Bio2Q

We are delighted to welcome Mr. Eshaan Mistry as a Research Internship Student of Bio2Q.

”Hello! My name is Eshaan Mistry, and I’m excited to introduce myself as an upcoming Bio2Q intern in Prof. Koyama’s group. I’m a third-year at UC Berkeley studying physics, computer science, and chemistry, though generally I’ve found that my interest has been continually developing through the deconstruction of complex systems. What counts as evidence, what are we assuming, and how do we move from data to understanding? A big theme in my academic journey has been learning how constraints shape what we can discover. Physics pulled me in because it trains you to reason from principles, and I quickly became fascinated by how seemingly “simple” facts introduce deep structure- how scale affects mass transport, energy behavior, and how physical laws must be interpreted carefully. As I moved through increasingly computation-heavy experiences, I also started to see how many research questions are really questions about representation: how do we encode something simply enough to learn from it without decomposing it to inanity? My first major research experience at NASA Ames reflected this shift toward computation and inference. The challenge was immediate: single-cell sequencing datasets are extremely high-dimensional, with practical bottlenecks to how much we can process. I learned how easy it is to accidentally choose constraints that limit the biology we can see and how critical it is to understand what’s lost when we reduce, denoise, cluster, or embed. As my interests evolved, I also became drawn to the way computation can convert unstructured knowledge into something actionable. I’ve worked on a machine-learning/NLP-style project aimed at extracting biochemical-relevant relationships from scientific literature. At Lawrence Berkeley National Laboratory, I explored machine-learning approaches for solid-state physics instrumentation optimization, including building surrogate models to speed up simulation workflows and improve performance through automated evaluation. Computation design proved repeatedly to be that which determines whether insight could be obtained efficiently enough to iteratively refine ideas into what we call ‘discovery’. What excites me most about this internship is the opportunity to work in a space of interdisciplinary expectation. I’m especially interested in working with researchers who can move fluidly between physics, machine learning, and biology, and who enjoy thinking at those boundaries. I’m also motivated by the chance to apply frontier computational ideas to real biological problems, such as metabolite prediction and quantum computation, while learning how to do this responsibly and rigorously with a team. Sparing any more technical verbiage, much of my time is spent outside of any lab! I enjoy practicing wushu (Chinese martial arts), reading (sci-fi , philosophy, poetry, critical works, anything really), origami, board games, listening to music, and just generally finding new ways of spending energy to expand my experience and understanding of everything around me. I’m very grateful for the opportunity to join Bio2Q. I truly look forward to learning from the group, contributing wherever I can, and growing alongside the team!’ “