by Chloe Canrong Qiu
Under the generous support of the Friends of Tokyo, Inc., I had the opportunity to participate in the UCEAP Summer Science Research Program this summer. I was accepted into Professor Kesuke Goda’s lab in the Department of Chemistry at the University of Tokyo, where I was assigned to Professor Walker Peterson’s team. While immersing myself in the vibrant life of Tokyo, I gained invaluable research experience at Goda Lab.
The Peterson Team was focused on developing and enhancing a cutting-edge imaging system they had invented – flow zoometry of Drosophila. This foundational technology platform enables high-throughput, large-scale imaging of Drosophila larvae. Despite my initial unfamiliarity with a biochemical lab setting, I soon found my role within the team as a Data Science major. Under Professor Peterson’s guidance, I spearheaded a project aimed at building an automated body orientation analyzer for Drosophila larvae. This project sought to improve the accuracy of downstream tumorigenic classification, where cancerous and non-cancerous larva samples were scanned using the flow zoometry system, highlighting the system’s significance and profound implications.
Breaking down the project into a series of manageable tasks, I first coded an Amira microscopy image visualization subprogram, which enables rapid production of 2D snapshots from 3D Drosophila larva image data. Using this subprogram, I processed over 3,100 Drosophila larva samples that our team had previously scanned, creating a dataset with body orientation annotations for the samples. Meanwhile, I identified and categorized some problematic scans as part of the data cleaning process. Based on the dataset I established, I experimented with various machine learning algorithms and developed a high-accuracy binary classifier using the random forest model, boosting downstream tumorigenic classification accuracy to 92%.
Research at UTokyo
It was a privilege to be exposed to and participate in real-world research projects at a globally renowned institution like the University of Tokyo. Leading my project within the Peterson Team allowed me to expand my knowledge and skills in data analysis and machine learning. During weekly meetings, I honed my communication skills by presenting in front of and learning from both scholars and students. I was particularly happy that my contributions extended beyond my project. For instance, my annotations of the Drosophila samples served as a standard reference for the team’s ongoing tumorigenic classification projects, and the subprogram I developed significantly reduced the image processing time for other team members working on their larva samples.
Beyond the laboratory, I explored and enjoyed Tokyo with friends from the UCEAP Program. From museums in Ueno to malls in Ikebukuro, from classic Edo-style sushi with Professor Goda and lab mates to the adorable bear onsen hot pot, from a solitary afternoon walk in the stylish Nakameguro neighborhood to a day trip in Kamakura visiting the Buddha and the seashore with friends… Under the fierce sunlight and occasional unexpected rains, we left our joy and laughter in the city and brought lasting memories home.
Life in Tokyo
In closing, I extend my heartfelt gratitude to the Friends of Tokyo, Inc., Professor Goda, and Professor Peterson for their unwavering support and guidance throughout this program. This experience has not only deepened my knowledge in data science and its application in cutting-edge research but also allowed me to forge meaningful friendships with fellow students and researchers. The insights and memories I have garnered during my time in Tokyo are undoubtedly precious, and I leave with a deep appreciation for both the academic and personal growth I have achieved.
