Cell analysis tools: an open-source library for single-cell analysis of multi-dimensional microscopy images
Emmanuel Contreras Guzman, Peter Rehani, Melissa C. Skala
Abstract
Single cell analysis of multi-dimensional microscopy images is repetitive, time consuming, and arduous. Numerous analysis steps are required to quantify and visualize cell heterogeneity and trends between experimental groups. The open-source community has created tools to facilitate this process. To further simplify analysis, we created a library of functions called cell-analysis-tools. This library includes functions that can streamline single-cell analysis for faster quality checking and automation. This library also includes example code with randomly generated data for dimensionality reduction [t-distributed stochastic neighbor embedding (t-SNE), principal component analysis (PCA), Uniform Manifold Approximation and Projection (UMAP)] and machine learning models [random forest, support vector machine (SVM), linear regression] that scientists can swap with their own data to visualize trends. Lastly, this library includes template scripts for feature extraction that can help identify differences between experimental groups and cell heterogeneity within a group. This library can significantly decrease user time while increasing robustness and reproducibility of results.