Journal: bioRxiv

Article Title: Data-driven simulations elucidate how lymphocyte motility behaviors drive cell-cell interactions within germinal centers

doi: 10.1101/2025.08.05.668700

Figure Lengend Snippet: From the intravital microscopy data, we segmented and tracked 3D trajectories for individual GCBs and Tfhs. Each trajectory comprised the (x,y,z) coordinates of a single cell in 20 consecutive frames acquired at 30-second intervals. For each experimental trajectory, we decomposed the trajectory into primary, secondary, and tertiary axes of movement. For the 3D trajectory, we quantified statistical features of the angular distribution. For the 3D trajectory and the trajectory decomposition, we quantified the following features: net distance and progressivity; statistical features of the displacement distribution; and the mean squared displacement at intervals of one, two, and three frames (30 seconds, 1-minute, and 1.5-minutes, respectively). After extracting these features, we projected the multi-dimensional feature space into a UMAP embedding and applied unsupervised clustering to identify motility behaviors. We then examined the behavior of each motility cluster. Finally, we eliminated possible outlier trajectories, and repeated step on the cleaned dataset.

Article Snippet: To extract single-cell trajectories from each intravital time-lapse microscopy image set, IMARIS software was used to segment cells and track individual trajectories.

Techniques: Intravital Microscopy

Journal: bioRxiv

Article Title: Data-driven simulations elucidate how lymphocyte motility behaviors drive cell-cell interactions within germinal centers

doi: 10.1101/2025.08.05.668700

Figure Lengend Snippet: A , Bar plot representing fractional volume of segmented and tracked cell types within each intravital microscopy experiment. Segmented cells refer to cells that were tracked for 5 continuous frames. Tracked cells refer to cells that were tracked for 20 continuous frames. B, Measured GCB-Tfh interactions in each intravital microscopy experiment. We measured interactions by processing the metric of volumetric overlap ratio between each tracked cell and any labelled cell type (which could have also been tracked for 20 frames or segmented for a minimum of 5 frames, see Methods ). C, Scatter plot of simulation predictions against experimental measurements for total Tfh interactions per GCB. Vertical errorbars represent the standard deviation from 100 simulation repetitions. Line indicates a linear regression with the intercept fixed at zero. D, Scatter plots of simulation predictions against experimental measurements for total GCB interactions per Tfh. Vertical errorbars represent the standard deviation from 100 simulation repetitions. Line indicates a linear regression with the intercept fixed at zero. E, Bar plot representing the log 2 fold change between simulation predictions and experimental measurements for total Tfh interactions per GCB (blue) or total Tfh interactions per GCB (red). ‘RMSE’: root mean squared error between average simulation predictions and experimental measurements, ‘r’: r-value for Spearman correlation, ‘p’: p -value for Spearman correlation.

Article Snippet: To extract single-cell trajectories from each intravital time-lapse microscopy image set, IMARIS software was used to segment cells and track individual trajectories.

Techniques: Intravital Microscopy, Standard Deviation