ALT Figure 1: Brain endothelial cells are enriched for the Notch ligands jag2b and dll4. A-D show cell clusters and ligand expression in single-cell RNA-seq data from 6-day-old zebrafish brains using t-SNE, dot and violin plots. E shows CellChat-predicted Notch interactions between endothelial cells and pericytes. F-G show green jag2b reporter signal with red brain endothelium in 4.5-day-old larvae; scale bars, 50 µm. By Huaxing Zi, Chi Dang, Weiyi Sima, Xiaoyu Li, Guozhu Ning, Li Sun, Qiang Wang and Jingjing Zhang. Source: https://www.biorxiv.org/content/10.64898/2026.10.05.756680v1 . CC BY 4.0: https://creativecommons.org/licenses/by/4.0/ . Complete Figure 1 extracted from the public PDF; surrounding page text and caption omitted. No scientific panels cropped or altered.

ALT Figure 1 shows the ROVER-dMRI reconstruction framework. Thick-slice MRI volumes acquired at rotated orientations are aligned to a common coordinate frame. A shared multi-resolution hash encoder and direction-specific neural-network outputs predict thin-slice signals. Slice-profile weights combine those predictions, which are fitted to measured data using a physics-informed loss. Figure 1: Qiang Liu et al. Source: https://www.biorxiv.org/content/10.64898/2026.09.25.754394v1. CC BY 4.0: https://creativecommons.org/licenses/by/4.0/. Complete figure downloaded from bioRxiv; no crop.

ALT Figure 1 shows the SFUMATO workflow: spatial gene-transcript coordinates are binned into local expression features, reduced with TruncatedSVD and clustered by a Bayesian Gaussian mixture model. A dendrogram merges related components, and posterior probabilities mix their colors to produce uncertainty-aware spatial transcript maps. Figure 1: Alessia Giustolisi, Christophe Avenel and Carolina Wählby. Source: https://www.biorxiv.org/content/10.64898/2026.09.25.754409v1. CC BY 4.0: https://creativecommons.org/licenses/by/4.0/. Complete figure downloaded from bioRxiv; no crop.

ALT Figure 1 diagrams animal behavior as a sequence of observed states and actions, a generative model with changing internal states, and inverse reinforcement learning that infers rewards and internal-state dynamics from actions. It shows how reward functions yield action values and policies that favor higher-value actions. Figure 1: Aditi Jha et al. Source: https://www.biorxiv.org/content/10.64898/2026.09.25.754554v1. CC BY 4.0: https://creativecommons.org/licenses/by/4.0/. Complete figure downloaded from bioRxiv; no crop.