Authors:

Yu Tin Lin

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

Imaging mass spectrometry enables spatially-resolved, label-free measurements of hundreds to thousands of compounds in tissue samples. Each imaging mass spectrometry experiment generates a dataset of hyperdimensional molecular images over the number of molecules and x, y coordinate values. Computational preprocessing and analysis are crucial to unravel interpretable biological patterns in the hyperdimensional molecular and pixel space of imaging mass spectrometry datasets. Herein, we describe a user-friendly, open-source data processing pipeline written in Python and R to streamline preprocessing, statistical analysis, statistical learning, and machine learning of big imaging mass spectrometry datasets. Functions include raw data import, baseline corrections, mass calibration, mass alignment, peak picking, peak integration, normalization, chemical database searching, diagnostic molecular pattern recognition, biochemical pathway/network analysis, image processing, volcano plot visualization, heatmap visualization, dimensionality reduction, clustering, image segmentation, and transfer learning. This user-friendly, open-source data processing package for imaging mass spectrometry enables researchers without programming and statistical background to access resources for mining big imaging data sets. Untargeted analysis via imaging mass spectrometry holds promise for discovery- based mapping of characteristic in situ molecular profiles in diseases of interest to empower mechanistic or targeted therapeutic efforts. 147

Source:

University of Florida / 2024

Topics:

No topics listed

Co-authors:

Yu Tin Lin