Search bioRxiv⌕ Search
Home/Archive/bioRxiv

bioRxiv · 10.1101/081364

Deep Learning for Imaging Flow Cytometry: Cell Cycle Analysis of Jurkat Cells

Eulenberg, P.·Koehler, N.·Blasi, T.·Filby, A.·Carpenter, A. E.·Rees, P.·Theis, F. J.·Wolf, F. A.
View original source ↗Export metadata ↓

Abstract

We show that deep convolutional neural networks combined with non-linear dimension reduction enable reconstructing biological processes based on raw image data. We demonstrate this by recon-structing the cell cycle of Jurkat cells and disease progression in diabetic retinopathy. In further analysis of Jurkat cells, we detect and separate a subpopulation of dead cells in an unsupervised manner and, in classifying discrete cell cycle stages, we reach a 6-fold reduction in error rate compared to a recent approach based on boosting on image features. In contrast to previous methods, deep learning based predictions are fast enough for on-the-fly analysis in an imaging flow cytometer.

Source connections

Explore related subjects

cell biologyPreprint

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Eulenberg, P., Koehler, N., Blasi, T., Filby, A., Carpenter, A. E., Rees, P., Theis, F. J., Wolf, F. A.. 2016-10-17. Deep Learning for Imaging Flow Cytometry: Cell Cycle Analysis of Jurkat Cells. https://doi.org/10.1101/081364

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

KEEP EXPLORING

Related preprints

bioRxiv2026

Spatiotemporal mapping of endothelial proliferative response in acute chemical liver injury

Endothelial cells are major contributors to liver injury response, regeneration, and chronic liver disease. Previous studies identified a proliferative endothelial response to acute liver injury; however, a comprehensive characterization of this adaptive response is missing. We have developed a novel pipeline for spatiotemporal analysis of proliferative response in the liver. We demonstrate that endothelial cells enter the cell cycle upon acute centrilobular injury with a delay compared to hepatocytes and in a spatially distinct pattern from hepatocytes. We show that endothelial cycling predominantly occurs in the pericentral zone, while hepatocyte cycling occurs uniformly throughout the lobule. Endothelial cycling in injured livers leads to heightened clonal expansion, and to serial division rarely observed in homeostatic livers. Consequently, injured lobules exhibit endothelial overabundance during regeneration. Our study provides an in-depth characterization of endothelial proliferative response to liver injury and establishes a novel pipeline for spatiotemporal mapping of hepatic cell states.

cell biology↗
bioRxiv2026

Pathological levels of α-synuclein increase basal mitophagy in Parkinson's dopamine neurons

Elevated -synuclein results in mitochondrial dysfunction, however the exact mechanism is poorly understood. -Synuclein localises to multiple subcellular sites including mitochondria, where it binds anionic lipids which regulate mitochondrial dynamics through DRP1. We hypothesised that pathogenic levels of -synuclein lead to dysregulation of mitophagy and lipid-dependent mitochondrial dynamics, causing mitochondrial and neuronal dysfunction. Using induced pluripotent stem cell-dopamine neurons (iPSC-DANs) carrying the -synuclein gene triplication mutation (SNCA-Trp) which causes an early-onset form of Parkinson's disease (PD), we established that elevated -synuclein increased basal mitophagy compared to healthy controls and overall loss of mitochondrial mass. This was associated with accumulation of basal mitophagy receptors BNIP3 and NIX and decreased proteasome activity, and was independent of PINK1-Parkin mitophagy. We identified reduced mitochondrial gene expression by single cell RNAseq, mirroring heightened mitophagy and mitochondrial depletion. We found increased levels of DRP1-dependent mitochondrial fission and perturbed levels of anionic lipids, including cardiolipin and phosphatidic acid, two lipids which directly bind to -synuclein. Pharmacological inhibition of DRP1 reversed the excessive basal mitophagy in SNCA-Trp neurons and restored deficits in cellular ATP and calcium release. Finally, we assessed neuronal function and found that impairment of mitochondrial dysfunction by elevated -synuclein led to reductions in neuronal firing activity. Inhibition of DRP1 rescued the neuronal activity deficit in SNCA-Trp neurons, suggesting that relieving - instead of increasing - mitophagy pressure may be beneficial in -synuclein-driven PD.

cell biology↗
bioRxiv2026

Effect of Varying Cigarette Smoke Concentrations and Durations on the Retinal Pigment Epithelium

Dry age-related macular degeneration (AMD) is a leading cause of blindness in the elderly, and treatment options are limited. Cigarette smoke (CS), the primary modifiable environmental risk factor for AMD, damages the retinal pigment epithelium (RPE) by inducing oxidative stress, which in turn activates multiple stress response pathways. Low levels of stress induce autophagy, a cell-recycling mechanism; moderate stress induces epithelial-mesenchymal transition (EMT); and high, sustained stress, once other cytoprotective pathways are exhausted, leads to cell death. All three pathways are implicated in AMD pathogenesis. Although CS is known to induce oxidative stress, its impact on these distinct stress response pathways remains understudied, and few studies have directly compared mouse and human RPE responses to the same stressor. Here, we examined the effect of varying concentrations and durations of cigarette smoke condensate (CSC) exposure on oxidative stress, autophagy, EMT, and cell death in ARPE-19 (human) and mouse RPE (MRPE) cells. Understanding which stress response pathways predominate at which CSC exposure levels may help guide the development of AMD treatments stratified by smoking history.

cell biology↗