Search bioRxiv⌕ Search
Home/Archive/bioRxiv

bioRxiv · 10.64898/2026.02.28.708664

A foundation AI model enhances electron microscopy image analysis

Du, M.·Wang, Y.·Xie, L.·Deng, G.·Guo, J.·Han, B.·Chen, Z.-H.·Rui, C.·Han, J.·Chen, Y.·Zhao, Y.·Cao, R.·Wang, F.·Li, K.·He, Y.·feng, x.
View original source ↗Export metadata ↓

Abstract

Electron microscopy (EM) and advanced volume EM have extensive applications in deciphering cellular ultrastructures for life sciences. However, quality issues of EM images significantly impede the precise analysis and discovery of nanoscale biological structures. Here, we present an unsupervised Foundation Model for Five Tasks (DF5T) model for image enhancement through denoising, deblurring, super-resolution, two-dimensional (2D) inpainting, and three-dimensional (3D) isotropic restoration. DF5T was trained on an extensive dataset of multi-source membrane-bound organelle EM images, consisting of a total of over 2.25 million images. On unseen data, DF5T outperformed the existing state-of-the-art models in all five tasks. Furthermore, DF5T substantially compensates for missing three-dimensional structural information and reveals significant ultrastructural alterations in organelle geometry following chemical treatment. We demonstrate that DF5T significantly enhances electron microscopy images quality by improving the accuracy of downstream organelle segmentation and 3D structural restoration, thereby contributing to future advances in biological research.

Explore related subjects

cell biologyPreprint

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Du, M., Wang, Y., Xie, L., Deng, G., Guo, J., Han, B., Chen, Z.-H., Rui, C., Han, J., Chen, Y., Zhao, Y., Cao, R., Wang, F., Li, K., He, Y., feng, x.. 2026-03-03. A foundation AI model enhances electron microscopy image analysis. https://doi.org/10.64898/2026.02.28.708664

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related preprints

bioRxiv2026

Functional characterization of Rho GTPase activating proteins SYDE1 and SYDE2

The human genome encodes more than 60 proteins containing Rho GTPase activating protein (RhoGAP) domains, many of which remain understudied with respect to their target specificity and biological roles. SYDE1 and SYDE2 are two such orphan RhoGAPs, for which there are few studies characterizing their biochemical and cellular functions and conflicting reports identifying their cognate GTPases. We previously identified SYDE1 and SYDE2 in a screen for substrates of the c-Jun N-terminal kinases. Here, we show that SYDE1 and SYDE2 are preferentially phosphorylated by JNK1 relative to other mitogen-activated protein kinases (MAPKs) at sites proximal to a kinase docking region. Purified SYDE1 and SYDE2 are shown to have significant catalytic GAP activity toward RhoA, Rac1, and Cdc42. However, neither up- nor down-regulation of SYDE1/2 expression leads to detectable changes in bulk GTP loading of any of these GTPases. Nevertheless, we demonstrate that SYDE1 and SYDE2, in a partially GAP-dependent manner, increase cell spreading and number of focal adhesions, and promote more directionally persistent migration in HEK293 cells. Together, these findings establish SYDE1 and SYDE2 as robust JNK substrates with catalytic activity toward a set of Rho GTPases and reveal basic functions of SYDE1 and SYDE2 in regulating cell morphology, adhesion, and migration.

cell biology↗
bioRxiv2026

The filopodial scaffold polyphosphate dictates cell adhesion-versus-invasion decisions

Inorganic polyphosphate (polyP) is an ancient polymer conserved across all life, serving cell type and location specific functions in every major compartment. Yet its role at the plasma membrane, where it accumulates to peak levels in many primary cells, is largely unknown. Here we identify polyP as a stabilizing component of filopodia, actin based membrane protrusions that govern cell adhesion, contact inhibition, and chemotaxis. Elevating cellular polyP increases filopodial stability and enhances cell adhesion, whereas reducing polyP accelerates filopodial disassembly and promotes cell migration. Mechanistically, we find that polyP acts as a structural filopodial scaffold, recruiting and organizing IRSp53, a membrane curvature inducing protein. We show that metastatic fibroblasts and breast cancer organoids carry markedly reduced and intracellularly reorganized polyP levels relative to their non transformed counterparts. Restoring endogenous polyP via lipid nanoparticle delivery suppresses their invasive phenotypes and reverses prometastatic gene expression signatures, implicating polyP as a primordial tumor suppressor.

cell biology↗
bioRxiv2026

Mitochondrial transfer mediates metabolic communication between beta cells and islet macrophages

Pancreatic islet macrophages support islet homeostasis and adapt their metabolic program in response to environmental cues, including beta cell released factors. Intercellular mitochondrial transfer is a biological process that modulates cellular responses. To test whether beta cells, which are strongly secretory, transfer mitochondria to islet macrophages, we generated mice with beta cell-specific expression of mitochondrial GFP (PhAMfloxIns1Cre). We demonstrate that beta cells transfer mitochondria to islet macrophages in vivo and in vitro. Diabetogenic stressors did not alter the frequency of mitochondrial transfer and macrophages containing beta cell-derived GFP exhibit increased protein synthesis rates. RNA-seq identified upregulation of activity-regulated cytoskeleton associated protein (Arc) in macrophages receiving beta cell-derived mitochondria, while disruption of actin cytoskeleton dynamics prevented mitochondrial transfer. Together, these findings identify mitochondrial transfer as a previously unrecognized mechanism of beta cell-macrophage communication that may contribute to islet homeostasis and immune regulation.

cell biology↗