Project Title: Biomarker discovery in metabolomics data for cervical cancer
Faculty Sponsor: Halil Bisgin
Department: Computer Science, Engineering, and Physics
Project Description: Cervical cancer (CC) is one of the most common causes of death from cancer in women. Even though there has been tremendous research to develop methods for early diagnosis of CC, further studies for biomarker discovery are still needed for consistent criteria. In addition to antibodies and mRNA, metabolic alterations that are specific to the development of stage offer opportunities for biomarker identification. In this project, we aim to discover metabolites that can predict a cancer case by employing machine learning tools. We hope to verify/discover metabolomics biomarkers in blood samples that can serve as early detection signals for CC.
Student Tasks & Responsibilities: Student will preprocess the metabolomics data and apply feature selection algorithms at the first place to discover the most useful metabolites for classification and clustering tasks.
Minimum Student Qualifications: The candidate should have exposure to statistics as well as programming. There's no preference on programming language as long as the student has necessary skills beyond 200-level CS programming courses.