Open-Source Analysis and Visualization Platform for Single-Cell MALDI-MSI Data

This project focuses on developing an open-source software platform to analyze and visualize Matrix Assisted Laser Desorption Ionization (MALDI) Mass Spectrometry Imaging (MSI) data. Tailored specifically for newly developed single-cell experiments, the software will process high-dimensional, noisy data that captures the molecular makeup of cells. By utilizing machine learning and a custom user interface, the platform will allow scientists to interact with the data and explore how different conditions such as drug treatments affect specific cell types.

Current software for analyzing MSI data is produced by large companies, closed-source, and increasingly expensive. Furthermore, these existing tools are built for general MSI experiments and lack the necessary functionality to handle novel single-cell experiments. Because single-cell MALDI-MSI data is incredibly dense, noisy, and high-dimensional, researchers currently struggle to effectively visualize the data, reduce its dimensionality, and identify the important features needed to verify scientific conclusions.

• Develop an Open-Source Platform: Create accessible, open-source software that bridges the gap left by expensive, proprietary MSI tools.
• Design an Interactive UI: Build a user interface capable of displaying dense, high-dimensional molecular data in a comprehensive and meaningful way.
• Implement Dimensionality Reduction: Apply unsupervised machine learning methods to reduce the high-dimensional spectra and represent the differences between cells.
• Map Data to Source: Ensure that the lower-dimension representations can be accurately connected back to their original single cells.
• Visualize Feature Importance: Provide tools to show feature importance, allowing scientists to connect features to experimental conditions and verify machine learning conclusions.
• Integrate Supervised Learning (Secondary): Include supervised learning methods to classify different cell types and conditions to make these tools more accessible to researchers.

• Proficient in python
• Good grasp of data structures
• Strong debugging abilities

• Experience creating user interfaces
• Background in biology/chemistry