Thomas Walter

The paper describes an optimized spatial transcriptomics workflow for Xenopus embryos, enabling high-resolution gene expression mapping across developmental stages.

HEDeST improves spatial transcriptomic deconvolution by integrating histology features, achieving single-cell resolution and outperforming existing methods in simulated and real cancer datasets.

RNA2seg is a cell segmentation model for spatial transcriptomics trained on MERFISH and CosMx data, showing superior performance in zero-shot and few-shot settings.

This paper introduces a contrastive learning and graph-based method to improve cell division detection and tracking in live-cell imaging data, achieving better performance across different temporal resolutions.

The paper predicts cell cycle phases from microscopy data using a variational auto-encoder with auxiliary tasks, and provides a new annotated dataset.

This paper develops a self-supervised learning model to predict iCCA transcriptomic classes from histological images, achieving good to very good performance in multiple validation sets.

This paper develops a deep learning model to predict molecular subtypes and outcomes in muscle-invasive bladder cancer using histopathological images instead of RNA sequencing.

Cut-Detector automates the analysis of cytokinesis timing from time-lapse microscopy movies by detecting microtubule cuts in intercellular bridges.

autoFISH provides a modular toolbox for automated smFISH experiments, including equipment plans, software, protocols, and analysis workflows validated on cell lines and tissues.

sCellST predicts single-cell gene expression from H&E images using spatial transcriptomics data, outperforming traditional methods in pancreatic cancer and detecting morphological differences in ovarian cancer.

This study examines how mRNA localization to P-bodies changes during the cell cycle, revealing that mRNA dynamics in P-bodies are not solely linked to translation status but are regulated by RNA features and RBPs.

This paper introduces ComSeg, a segmentation algorithm for image-based spatial transcriptomics that outperforms existing methods in cell type calling and handles complex tissue structures.

This paper introduces Giga-SSL, a self-supervised method for learning Whole Slide Image representations from The Cancer Genome Atlas data, improving classification performance and enabling efficient analysis.

pyHiM provides a modular, open-source software package for analyzing spatial genomics data from multiplexed DNA-FISH imaging, enabling chromatin trace reconstruction and data validation.

This paper presents a segmentation workflow that enables cell segmentation without structural fluorescent labels, improving performance and robustness across cell lines and fluorescent labels.

This study develops a deep learning model to classify head and neck squamous lesions with confidence scores to identify uncertain predictions.

The paper proposes a modified cross-validation approach and a new neural network architecture for treatment response prediction in small Triple Negative Breast Cancer cohorts.

FISH-quant v2 is a modular tool for analyzing smFISH images, enabling detection, segmentation, quantification, and visualization of RNA localization at the single-cell level.

The paper develops a fast image classification method for biological microscopy by selecting discriminant features and using machine learning to improve speed and accuracy.

The paper predicts drug mechanism of action across multiple cell lines using domain-invariant features from multitask autoencoders.

This paper introduces kernel multitask regression to predict compound toxicity on cell lines using chemical and genetic data, achieving strong performance on a benchmark challenge.