Susanne Gerber

This study compares machine learning and statistical methods for detecting polygenic adaptation using simulated evolve-and-resequence data, finding that the OCSVM-FET approach performs best.

DModE provides an end-to-end framework for analyzing Nanopore DRS data, enabling integrated differential expression and modification analysis with automated reporting.

This study investigates how 16p11.2 copy number variations affect genome architecture and transcription during neurodevelopment, finding both local dosage effects and broader chromatin changes.

The paper benchmarks ambient RNA removal tools across different platforms, finding that scAR and CellClear generate artificial counts and cell types, while CellBender and SoupX perform reliably.

The paper identifies preprocessing bias in chromatin contact data and introduces a framework to correct it, enabling physically interpretable genome architecture reconstruction.

This study introduces NanoRibolyzer, a long-read nanopore sequencing method to map pre-rRNA processing and modification at single-nucleotide resolution, revealing cleavage sites and modification patterns.

This paper demonstrates real-time transcriptomic analysis using nanopore sequencing for quality control and differential expression in various experimental conditions.

This study evaluates improvements in direct RNA sequencing (DRS) using RNA004 chemistry and modified-base-calling models, showing enhanced throughput, accuracy, and detection of RNA modifications in clinical applications.

ModiDeC is a deep-learning classifier that identifies multiple RNA modifications from direct nanopore sequencing, validated on synthetic data and biological samples.

The paper introduces DLRN, a deep learning framework that models time-resolved data to extract kinetic reaction networks, time constants, and amplitudes with performance comparable to or better than classical methods.

ClearFinder provides a Python GUI for cell annotation in cleared mouse brains, improving accessibility and enabling statistical analysis and visualization.

The paper analyzes miRNA expression and co-targeting relationships during mouse cerebral cortex development, identifying stage-specific hubs and enhanced transcriptional silencing effects of miRNA pairs.

This study identifies miR-92a-3p as a regulator of RBFOX2 expression, which controls the splicing transition from neural progenitors to neurons in the developing mouse cortex.

Cortexa provides a web-portal with standardized RNA-sequencing data and tools for analyzing gene expression and alternative splicing in the mouse brain.

Neuroimage Denoiser removes noise from functional microscopy recordings, improving neural activity analysis by preserving signal amplitude and operating without re-training for different sensors.

Direct CCUT trains deep models to reconstruct high-resolution chromatin interaction matrices from low-resolution Pore-C and 3C data using consumer GPUs.

The paper identifies sustained freezing during prolonged threat memory retrieval as an anxiety-endophenotype biomarker in both male and female inbred mice, with sex-specific transcriptomic differences in brain regions associated with psychiatric disorders.

The paper shows that m1A methylation of ND5 mRNA, catalyzed by TRMT10C, leads to translation repression and mitochondrial dysfunction in Alzheimer's disease.

The paper demonstrates that ignoring sampling time points in statistical analyses can lead to misinterpretation of interaction effects between biomarkers and cognitive reserve in Alzheimer's Disease.

NanopoReaTA provides a user-friendly tool for real-time RNA-seq analysis of nanopore sequencing data, enabling interactive decision-making during sequencing.

The paper shows that reactivation of inactive X-chromosome alleles in female neural cells during differentiation enhances resilience to X-linked neurodevelopmental disorders.

This study investigates the lifelong molecular effects of high glucocorticoid exposure during development using a zebrafish model, identifying GC-primed genes and epigenetic changes linked to psychiatric disorder risk factors.

The paper introduces a probabilistic Mumford-Shah denoising model that effectively reduces noise in 3D images and 2D videos while maintaining quality and computational efficiency.

The paper introduces a new file format (tbg) for genomic data and associated tools to efficiently analyze SNPs and predict amino acid changes.

This study identifies mTORC1 as a transcriptional regulator of cholesterol biosynthesis genes in developing neurons, linking mTOR activity to cholesterol production and neurodevelopmental disorders.

The paper introduces a low-cost computational tool for discretization, prediction, and feature selection in complex systems, achieving significant cost and performance improvements.

The paper introduces a new relation measure, Linkage Probability, to analyze genetic data and reveals 3D chromatin loop effects in SNP data.