Publication Date
Fall 2025
Degree Type
Master's Project
Degree Name
Master of Science in Bioinformatics (MSBI)
Department
Computer Science
First Advisor
Wendy Lee
Second Advisor
Philip Heller
Third Advisor
Fabio Di Troia
Keywords
TransUNET, Segmentation, Gel Electrophoresis, NextFlow
Abstract
Molecular biology experiments traditionally produce nucleic acid or protein products. When an experiment produces these molecular products, the first step in determining the success is to validate the products. To accomplish this for nucleic acids, a charge separation-based technique called gel electrophoresis is used. The result is a gel that can be imaged and analyzed by the combination of specialized tools and scientific expertise. Given the potential expense and difficulty associated with this analysis, computer vision can be utilized to generate a more efficient pipeline. This project aims to adapt an image segmentation technology framework called TransUnet to validate and analyze the results of nucleic acid experiments. Specifically, the goals of this project are to identify bands in gel electrophoresis. The key components of this pipeline include data acquisition, pre-processing, data validation, establishing training/testing sets, and concentration calculation.
Recommended Citation
Kalampukattussery, Allen, "The Implementation of TransUNET Infrastructure for Gel Band Image Segmentation" (2025). Master's Projects. 1588.
https://scholarworks.sjsu.edu/etd_projects/1588
