Search GIDE databases

Original images

Imaging Methods: bright-field microscopy confocal microscopy

Organisms: Platynereis dumerilii

Figure 6 Raw Data

Organisms: Setaria viridis

The subcellular localization of a protein is important for its function, and its mislocalization is linked to numerous diseases. Existing datasets capture limited pairs of proteins and cell lines, and existing protein localization prediction models either miss cell-type specificity or cannot generalize to unseen proteins. Here we present a method for Prediction of Unseen Proteinsí Subcellular localization (PUPS). PUPS combines a protein language model and an image inpainting model to utilize both protein sequence and cellular images. We demonstrate that the protein sequence input enables generalization to unseen proteins, and the cellular image input captures single-cell variability, enabling cell-type-specific predictions. Experimental validation shows that PUPS can predict protein localization in newly performed experiments outside of the Human Protein Atlas used for training. Collectively, PUPS provides a framework for predicting differential protein localization across cell lines and single cells within a cell line, including changes in protein localization driven by mutations.

Imaging Methods: spinning disk confocal microscopy

Organisms: Homo sapiens

The mechanisms underlying early Parkinson’s disease (PD) remain unexplained. While the aggregation of alpha-synuclein (αSyn) into Lewy bodies (LBs) characterises the pathology of later stages, emerging evidence suggests that small, toxic αSyn oligomers may drive disease during the pre-symptomatic phase. Here, we introduce the concept of Aggregation-Susceptible Cells (ASCs): neurons that, due to elevated intracellular αSyn concentration, are predisposed to forming pathological aggregates. Using quantitative imaging, we identified 9,882 neurons containing 112 million αSyn oligomers across multiple brain regions from post-mortem human tissue of early-stage PD cases and controls. No increase in intracellular αSyn concentration in early disease was found; in other words, there is no “unique cell” in PD with an unexpectedly high αSyn oligomer concentration. However, the proportion of ASCs was significantly elevated in brain regions undergoing early pathological manifestation. This finding supports a model in which idiopathic PD pathogenesis is not driven by a change in αSyn aggregation kinetics at the protein level, but rather by an increased prevalence of neurons stochastically crossing a critical concentration threshold for aggregation at the cell-population level. These results identify cell aggregation susceptibility as a fundamental mechanism in the earliest transition to pathology and offer a new quantitative framework for understanding early-stage protein misfolding in PD.

Imaging Methods: spinning disk confocal microscopy

Organisms: Homo sapiens

Research utilising exercise therapeutically to enhance neuroplasticity mainly reports enhanced expression of neurotrophic (neural growth) factors which are associated with increased neurogenesis, structural changes, and improved memory. However, these neural changes are tightly regulated by both promotors and inhibitors of neuroplasticity, and the effect of exercise on inhibitory pathways is currently understudied. We show that exercise also modulates inhibitors of neuroplasticity. Six weeks of treadmill training reduced the gene expression of aggrecan, a chondroitin sulphate proteoglycan (CSPG), and altered the composition of CSPG containing structures, perineuronal nets, in the rodent hippocampus. Genetically manipulating chondroitin-4-sulphation of CSPGs by overexpressing hippocampal Chst11 impaired memory performance, which was mitigated by treadmill training. We provide evidence that hippocampal CSPGs are involved in object recognition memory, and that there is a link between exercise, modulating hippocampal inhibitors of neuroplasticity, and memory performance.

Imaging Methods: fluorescence microscopy

Organisms: Rattus norvegicus

This dataset comprises live-cell fluorescence resonance energy transfer (FRET) and corresponding bright-field images of MCF-7 cells co-expressing the fluorescently tagged proteins CFP-BCL-XL and YFP-BAK under various drug treatment conditions. The data were acquired using quantitative FRET microscopy to simultaneously capture dynamic drug-target interactions and morphological changes at single-cell resolution. This resource enables the quantitative analysis of drug-induced modulation of BCL-XL/BAK interactions, the correlation between target engagement and phenotypic outcomes, and the development of computational models for drug efficacy scoring. The dataset is an essential resource for researchers in cancer biology, drug discovery, and high-content image analysis, providing a foundation for studying dynamic protein-protein interactions and functional drug responses in a live-cell context.

Organisms: Homo sapiens

This dataset comprises live-cell fluorescence resonance energy transfer (FRET) and corresponding bright-field images of MCF-7 cells co-expressing the fluorescently tagged proteins CFP-BCL-2 and YFP-BAK under various drug treatment conditions. The data were acquired using quantitative FRET microscopy to simultaneously capture dynamic drug-target interactions and morphological changes at single-cell resolution. This resource enables the quantitative analysis of drug-induced modulation of BCL-2/BAK interactions, the correlation between target engagement and phenotypic outcomes, and the development of computational models for drug efficacy scoring. The dataset is an essential resource for researchers in cancer biology, drug discovery, and high-content image analysis, providing a foundation for studying dynamic protein-protein interactions and functional drug responses in a live-cell context.

Organisms: Mus musculus

This dataset comprises live-cell fluorescence resonance energy transfer (FRET) and corresponding bright-field images of H1975 cells co-expressing the fluorescently tagged proteins CFP-EGFR and YFP-GRB2 under various drug treatment conditions. The data were acquired using quantitative FRET microscopy to simultaneously capture dynamic drug-target interactions and morphological changes at single-cell resolution. This resource enables the quantitative analysis of drug-induced modulation of EGFR/GRB2 interactions, the correlation between target engagement and phenotypic outcomes, and the development of computational models for drug efficacy scoring. The dataset is an essential resource for researchers in cancer biology, drug discovery, and high-content image analysis, providing a foundation for studying dynamic protein-protein interactions and functional drug responses in a live-cell context.

Organisms: Homo sapiens

This dataset comprises live-cell fluorescence resonance energy transfer (FRET) and corresponding bright-field images of A549 cells co-expressing the fluorescently tagged proteins CFP-EGFR and YFP-GRB2 under various drug treatment conditions. The data were acquired using quantitative FRET microscopy to simultaneously capture dynamic drug-target interactions and morphological changes at single-cell resolution. This resource enables the quantitative analysis of drug-induced modulation of EGFR/GRB2 interactions, the correlation between target engagement and phenotypic outcomes, and the development of computational models for drug efficacy scoring. The dataset is an essential resource for researchers in cancer biology, drug discovery, and high-content image analysis, providing a foundation for studying dynamic protein-protein interactions and functional drug responses in a live-cell context.

Organisms: Homo sapiens

A time series of Heterodera schachtii infecting MAGIC population of Arabidopsis thaliana.

Organisms: Arabidopsis thaliana

Standard RGB and multi-spectral images generated from VideometerLab4 for three wheat NDM lines' seeds

Organisms: Triticum aestivum

This collection includes 3D FIB-SEM datasets of Phaeodactylum tricornutum (strain 1 and strain 18) with segmentation labels for chloroplast and vacuole. It provides pretrained deep learning models developed across multiple architectures (3D VNet, 3D SegNet, 3D CNN 4-level, MONAI DynUNet, MONAI BasicUNet, Swin-Net + VNet hybrid, and Contrast-Aware Dual Encoder VNet). The dataset supports organelle segmentation research and transfer learning.

Imaging Methods: focussed ion beam scanning electron microscopy (FIB-SEM)