Geisberger, R. | Search bioRxiv

bioRxiv2025

spatialgater: A shiny webtool for in situ gating of cells in SpatialExperiments

Multiplexed imaging techniques generate high-dimensional datasets that contain their molecular profiles of cells combined with spatial coordinates, which can be stored in SpatialExperiment objects. Current analysis workflows separate cells after clustering them by their bio-molecule expression levels without considering their spatial context within the tissue. While patch-/neighbourhood detection methods exist, there is no option to select single cells by their location at will. By introducing spatialgater, we aim to boost interactivity and reduce programming efforts of image analysis by enabling spatial gating of cells from SpatialExperiment objects through an intuitive web-based interface. The package displays cells as dots on a zoom-able image and allows users to draw polygon gates directly on individual cells. All selected cell identifiers can be exported as.csv and/or written back as logical gate column in the colData() of the original SpatialExperiment. spatialgater provides an accessible, interactive addition to spatial analysis pipelines. By using a publicly available imaging mass cytometry dataset from breast cancer tissue, we demon-strate its effectiveness in characterizing and comparing T-cells by their spatial location.

bioinformatics↗

bioRxiv2022

Validation of genetic variants from NGS data using Deep Convolutional Neural Networks

Accurate somatic variant calling from next-generation sequencing data is one most important tasks in personalised cancer therapy. The sophistication of the available technologies is ever-increasing, yet, manual candidate refinement is still a necessary step in state-of-the-art processing pipelines. This limits reproducibility and introduces a bottleneck with respect to scalability. We demonstrate that the validation of genetic variants can be improved using a machine learning approach resting on a Convolutional Neural Network, trained using existing human annotation. In contrast to existing approaches, we introduce a way in which contextual data from sequencing tracks can be included into the automated assessment. A rigorous evaluation shows that the resulting model is robust and performs on par with trained researchers following published standard operating procedure.

bioinformatics↗