Manfred Claassen

Mechanistic gene-regulation-based trajectory inference with bowside

The paper introduces bowside, a neural model for trajectory inference and gene regulatory network reconstruction in single-cell RNA sequencing data, showing it recovers known regulatory motifs in hematopoiesis and reprogramming.

Accurate quantification of spliced and unspliced transcripts for single-cell RNA sequencing with tidesurf

The paper introduces tidesurf, a tool for quantifying spliced and unspliced transcripts in scRNA-seq, showing improved accuracy over velocyto for 5' chemistry data.

G-LATO: Inference of Spatial Latent Ordering via Deep Gaussian Processes

G-LATO infers spatial gradients using deep Gaussian processes, revealing known and novel spatial organization in liver and glioblastoma data.

t2pmhc: A Structure-Informed Graph Neural Network to predict TCR-pMHC Binding

The paper introduces t2pmhc, a structure-informed graph neural network for predicting TCR-pMHC binding, showing improved generalization to unseen peptides compared to existing methods.

velotest: Statistical assessment of RNA velocity embeddings reveals quality differences for reliable trajectory visualizations

velotest assesses RNA velocity embeddings' faithfulness to high-dimensional data, revealing quality differences across methods and datasets.

Quantification of spliced and unspliced transcripts by velocyto is inaccurate for 5'-sequencing data

This study shows that velocyto inaccurately quantifies spliced and unspliced transcripts for 5'-sequencing data, leading to incorrect velocity predictions.

Factorial state-space modelling for kinetic clustering and lineage inference

This paper introduces a latent state-space Markov model for kinetic clustering and lineage inference in scRNA-seq data, demonstrating its ability to distinguish functional from transitional cell-states during differentiation.

S3-CIMA: Supervised spatial single-cell image analysis for the identification of disease-associated cell type compositions in tissue

Enhanced endogenous activity predicts elimination of adult-born neurons in the mouse olfactory bulb

The paper identifies enhanced endogenous Ca2+ signaling as a predictor of elimination in adult-born neurons in the mouse olfactory bulb.

Phase-specific signatures of wound fibroblasts and matrix patterns define cancer-associated fibroblast subtypes

The paper identifies wound fibroblast phases and matrix patterns that define cancer-associated fibroblast subtypes with prognostic relevance in skin cancer.

Cytopath: Simulation based inference of differentiation trajectories from RNA velocity fields

Cytopath infers differentiation trajectories using RNA velocity data, outperforming existing methods in reconstructing complex differentiation processes.

Positive feedback induces switch between distributive and processive phosphorylation of Hog1

The paper investigates the phosphorylation mechanism of Hog1 in yeast, finding that positive feedback enables a switch between distributive and processive phosphorylation, enhancing sensitivity and robustness.

Dynamic Distribution Decomposition for Single-Cell Snapshot Time Series Identifies Subpopulations and Trajectories during iPSC Reprogramming

DDD identifies dynamically important cell states and trajectories during iPSC reprogramming by modeling single-cell time series as a continuous Markov process.

psupertime: supervised pseudotime inference for single cell RNA-seq data with sequential labels

psupertime improves pseudotime inference in single-cell RNA-seq by using sequential labels to identify genes relevant to biological processes rather than other sources of variation.

Automated Gleason grading of prostate cancer tissue microarrays via deep learning

This study develops a deep learning model to automate Gleason grading of prostate cancer tissue microarrays, achieving agreement with pathologists and expert-level patient stratification.

Supervised learning on synthetic data for reverse engineering gene regulatory networks from experimental time-series

This paper introduces a supervised learning method using synthetic data to reconstruct gene regulatory networks from experimental time-series, showing superior performance compared to unsupervised methods.

Tree-ensemble analysis tests for presence of multifurcations in single cell data

TreeTop identifies and assesses branch points in single-cell data, showing shallow differentiation hierarchies in hematopoiesis.

Sensitive detection of rare disease-associated cell subsets via representation learning.

This paper introduces CellCnn, a representation learning method to detect rare disease-associated cell subsets from single-cell data, identifying subsets with frequencies as low as 0.005%.