Carsten Marr

ExPert infers causal perturbations from cell states across biological and clinical contexts

ExPert infers causal perturbations from cell states by learning prototypes across 12 contexts, achieving high recall with low compute and validating findings in disease models.

BaSiCPy: Scalable and Robust Shading Correction for Optical Microscopy Images

BaSiCPy improves illumination correction in microscopy images by addressing foreground correlations and enabling automated processing.

A CAR T cell-amplified suppressive B cell circuit in AML

The paper identifies suppressive atypical memory B cells in AML that are amplified by CAR T cell interactions and shows that targeting both B cells and AML blasts can overcome this suppression.

TopoMIL: Topology Improves Multiple Instance Learning in Diagnostic Microscopic Images

TopoMIL improves multiple instance learning for diagnostic microscopic images by integrating topological structures, achieving classification gains across various pooling methods.

Leveraging Twin Information for Inference of Gene Regulatory Networks

This paper introduces TwINFER, a framework using twin cell information to improve gene regulatory network inference, resolving directionality and reducing false positives in network motifs.

EchoVisuALL: From Echocardiography to Gene Discovery

EchoVisuALL automates echocardiography analysis in mice, identifying 37 genes linked to cardiac abnormalities through phenotyping and clustering.

Live high-content imaging with automated analysis reveals mitochondrial changes during vascular calcification

This paper introduces H-MIP, a high-throughput platform for mitochondrial morphology analysis, and applies it to study mitochondrial changes in vascular calcification.

Leonardo: a toolset to correct sample-induced artifacts in light sheet microscopy images

Leonardo corrects sample-induced artifacts in light sheet microscopy images using two modules, validated on diverse biological samples.

Predictive value of preclinical models for CAR-T cell therapy clinical trials: a systematic review and meta-analysis

This study reviews and meta-analyzes CAR-T cell therapy trials and preclinical models to assess their predictive value, finding that preclinical models poorly predict outcomes when using their own features.

Single-cell multiome uncovers differences in glycogen metabolism underlying species-specific speed of development

The paper identifies glycogen metabolism differences in pluripotent cells across species that influence neural differentiation speed.

A scalable computational framework for annotation-independent combinatorial target identification in scRNA-seq databases

The paper introduces otopia, a framework for querying large scRNA-seq databases to identify cell populations matching complex gene expression patterns without relying on annotations.

Cytosolic S100A8/A9 promotes Ca2+ supply at LFA-1 adhesion clusters during neutrophil recruitment

The paper shows that cytosolic S100A8/A9 is essential for neutrophil adhesion under flow by regulating Ca2+ availability at LFA-1 adhesion clusters.

Spatial proteomics in neurons at single-protein resolution

This paper introduces SUM-PAINT, a high-throughput imaging method for single-protein spatial resolution, and discovers a new synapse type through extensive multiprotein analysis.

scifAI: Explainable machine learning for profiling the immunological synapse and functional characterization of therapeutic antibodies

scifAI provides an open-source framework for analyzing imaging flow cytometry data to predict antibody function and characterize immunological synapses.

Label-free imaging of 3D pluripotent stem cell differentiation dynamics on chip

The paper introduces a label-free deep learning method to predict nuclear staining in 3D stem cell cultures from bright-field images, enabling high-resolution tracking of differentiation dynamics.

PXPermute: Unveiling staining importance in multichannel fluorescence microscopy

PXPermute identifies important fluorescent channels in IFC data by permuting pixel values and evaluating model performance, showing superior accuracy in channel selection.

A generalizable brain extraction net (BEN) for multimodal MRI data from rodents, nonhuman primates, and humans

The paper introduces BEN, a deep learning method for brain tissue extraction across species and MRI modalities, showing improved accuracy and generalizability over existing tools.

Gal1 repression memory in budding yeast

This study quantifies repression memory in budding yeast cells during repeated carbon source shifts, showing shortened repression delays and a gain-of-memory phenotype in Elp6-deficient cells.

SHAPR predicts 3D cell shapes from 2D microscopic images

SHAPR predicts 3D cell shapes from 2D images, improving accuracy over existing methods and enabling morphometric analysis of cell aggregates.

InstantDL - An easy-to-use deep learning pipeline for image segmentation and classification

InstantDL provides a deep learning pipeline for biomedical image tasks with minimal effort, achieving competitive performance across seven datasets without parameter tuning.

Systematic comparison of incomplete-supervision approaches for biomedical imaging classification

The paper compares incomplete-supervision methods for biomedical image classification, finding that pre-training and semi-supervised learning contribute more than active learning to performance.

Computational modeling reveals cell-cycle dependent kinetics of H4K20 methylation states during Xenopus embryogenesis

The paper models H4K20 methylation kinetics during Xenopus embryogenesis, showing demethylation is essential in non-cycling cells but dispensable in rapidly dividing cells.

Gene networks with transcriptional bursting recapitulate rare transient coordinated expression states in cancer

The paper shows that transcriptional bursting in gene networks can produce rare coordinated high expression states in cancer cells, which are linked to drug resistance.

Aggregated spatio-temporal division patterns emerge from reoccurring divisions of neural stem cells

The paper shows that aggregated spatio-temporal division patterns in zebrafish neural stem cells arise from re-dividing cells, not random activity or population feedback.

Automatic identification of relevant genes from low-dimensional embeddings of single cell RNAseq data

This paper introduces methods to identify globally and locally relevant genes in nonlinear low-dimensional embeddings of single-cell RNA-seq data, demonstrating their versatility across datasets and embedding techniques.

Domain model explains propagation dynamics and stability of K27 and K36 methylation landscapes

The paper models the propagation and stability of K27 and K36 methylation landscapes using computational methods and mass spectrometry data.

Evaluation of domain adaptation approaches for robust classification of heterogeneous biological data sets

This study evaluates domain adaptation methods to improve classification accuracy in heterogeneous biological data, finding that EasyAdapt outperforms other techniques.

Subtle Changes in Clonal Dynamics Underlie the Age-Related Decline in Neurogenesis

This study identifies an increase in asymmetric stem cell divisions with age as a key factor in the decline of neurogenesis in adult mice.

cgCorrect: A method to correct for confounding cell-cell variation due to cell growth in single-cell transcriptomics

cgCorrect corrects cell size-related variability in single-cell transcriptomics data, improving analysis of gene expression heterogeneity.

destiny – diffusion maps for large-scale single-cell data in R

The paper introduces destiny, an R package for diffusion maps in single-cell data, with efficient nearest-neighbour approximation and noise modeling for large datasets.