Kleino Iivari, Perk Mats, Sousa António G G, Linden Markus, Mathlin Julia, Giesel Daniel, Frolovaite Paulina, Pietilä Sami, Junttila Sini, Suomi Tomi, Elo Laura L
Turku Bioscience Centre, University of Turku and Åbo Akademi University, Turku, FI-20520, Finland.
Institute of Biomedicine, University of Turku, Turku, FI-20014, Finland.
Bioinform Adv. 2025 Apr 4;5(1):vbaf069. doi: 10.1093/bioadv/vbaf069. eCollection 2025.
The analysis of cell migration using time-lapse microscopy typically focuses on track characteristics for classification and statistical evaluation of migration behaviour. However, considerable heterogeneity can be seen in cell morphology and microscope signal intensity features within the migrating cell populations.
To utilize this information in cell migration analysis, we introduce here an R package CellRomeR, designed for the phenotypic clustering of cells based on their morphological and motility features from microscopy images. Utilizing machine learning techniques and building on an iterative clustering projection method, CellRomeR offers a new approach to identify heterogeneity in cell populations. The clustering of cells along the migration tracks allows association of distinct cellular phenotypes with different cell migration types and detection of migration patterns associated with stable and unstable cell phenotypes. The user-friendly interface of CellRomeR and multiple visualization options facilitate an in-depth understanding of cellular behaviour, addressing previous challenges in clustering cell trajectories using microscope cell tracking data.
CellRomeR is available as an R package from https://github.com/elolab/CellRomeR.
使用延时显微镜分析细胞迁移通常侧重于轨迹特征，以对迁移行为进行分类和统计评估。然而，在迁移细胞群体中，细胞形态和显微镜信号强度特征存在相当大的异质性。
为了在细胞迁移分析中利用这些信息，我们在此引入一个R包CellRomeR，它旨在根据显微镜图像中细胞的形态和运动特征对细胞进行表型聚类。利用机器学习技术并基于迭代聚类投影方法，CellRomeR提供了一种识别细胞群体异质性的新方法。沿着迁移轨迹对细胞进行聚类，可以将不同的细胞表型与不同的细胞迁移类型相关联，并检测与稳定和不稳定细胞表型相关的迁移模式。CellRomeR用户友好的界面和多种可视化选项有助于深入了解细胞行为，解决了以往使用显微镜细胞跟踪数据对细胞轨迹进行聚类时遇到的挑战。
CellRomeR作为一个R包可从https://github.com/elolab/CellRomeR获取。
