bioRxiv Science⌕ Search

Biology subjects

Duerr, A.

Publications and source records attributed to Duerr, A..

2 recordsLinked to original sources

Single-cell immune profiling reveals the impact of antiretroviral therapy on HIV-1-induced immune dysfunction, T cell clonal expansion, and HIV-1 persistence in vivo

Despite antiretroviral therapy (ART), HIV-1 persists in proliferating T cell clones that increase over time. To understand whether early ART affects HIV-1 persistence in vivo, we performed single-cell ECCITE-seq and profiled 89,279 CD4+ T cells in paired samples during viremia and after immediate versus delayed ART in six people in the randomized interventional Sabes study. We found that immediate ART partially reverted TNF responses while delayed ART did not. Antigen and TNF responses persisted despite immediate ART and shaped the transcriptional landscape of CD4+ T cells, HIV-1 RNA+ cells, and T cell clones harboring them (cloneHIV-1). Some HIV-1 RNA+ cells reside in the most clonally expanded cytotoxic T cell populations (GZMB and GZMK Th1 cells). CloneHIV-1+ were larger in clone size, persisted despite ART, and exhibited transcriptional signatures of antigen, cytotoxic effector, and cytokine responses. Using machine-learning algorithms, we identified markers for HIV-1 RNA+ cells and cloneHIV-1+ as potential therapeutic targets. Overall, by combining single-cell immune profiling and T cell expansion dynamics tracking, we identified drivers of HIV-1 persistence in vivo.

immunology↗

DeepFLaSh, a deep learning pipeline for segmentation of fluorescent labels in microscopy images

Fluorescent labeling of biomolecules is widely used for bioimage analyses throughout the life sciences. Recent advances in deep learning (DL) have opened new possibilities to scale the image analysis processes through automation. However, the annotation of fluorescent features with a low signal-to-noise ratio is frequently based on subjective criteria. Training on subjective annotations may ultimately lead to biased DL models yielding irreproducible results. An end-to-end analysis process that integrates data annotation, ground truth estimation, and model training can mitigate this risk. To highlight the importance of this integrated process, we compare different DL-based analysis approaches. Based on data from different laboratories, we show that ground truth estimation from multiple human annotators is indispensable to establish objectivity in fluorescent feature annotations. We demonstrate that ensembles of multiple models trained on the estimated ground truth establish reliability and validity. Our research provides guidelines for reproducible and transparent bioimage analyses using DL methods.

neuroscience↗