by Sarah Bender
At the University Libraries, the Data and Code Support service offers specialized, one-on-one assistance for students, faculty, and staff using open source coding languages and data science tools in their research or projects. Whether you're wrangling a massive dataset or learning to code for the first time, CMU Libraries’ Data & Code Support team offers personalized, discipline-specific guidance to empower your research and accelerate discovery.
Systems Neuroscience Ph.D. student Isabella Salas-Allende was just beginning her work with an alpha-synuclein mouse model, used to study pathology and behavior associated with Parkinson's disease, when she reached out to Research Data Services Librarian Alfredo González-Espinoza for help. She wanted to use the position of the mouse’s nose to quantify tremor in the model, and she asked the Data and Code consultant for help understanding the theoretical and practical aspects of tremor analysis using frequency-domain methods.
“Dr. Gonzalez-Espinoza was a huge help in getting me started,” Salas-Allende recalled. “His support was incredibly thorough and patient, helping me build a solid foundation for my analysis. I highly recommend students to reach out!”
Goal
- To interpret data and quantify tremor in an alpha-synuclein mouse model, compared to littermate controls.
- To overcome roadblocks that arose over the course of the research.
How We Helped
- Salas-Allende collected data using DeepLabCut, an open-source Python package for animal pose estimation, to track the nose-point position of her mouse model. Gonzalez-Espinoza then helped her apply the fundamentals of the Fourier Transform — which is used to isolate and examine different frequencies — to the data. Using this method, she could better understand the experimental data by revealing key periodic features over time.
- They also explored the concept of power spectrum, which measures the energy of a signal at each frequency component. Gonzalez-Espinoza provided resources like videos and a custom simulation to tie these concepts to Salas-Allende’s research. Finally, he helped Salas-Allende interpret power spectral density in the context of her data. This provided her with a robust framework to analyze data from future experiments in her PhD research.
Results
- With advice and resources from Gonzalez-Espinoza, Salas-Allende was able to understand her code and interpret the data analysis results for her team’s new model.
- She is currently working on preparing a manuscript for submission.