Projects

Arrhythmogenic Cardiomyopathy (ACM)

Headed by Camilla Schinner and the Schinner Lab (Department for BioMedical Research - DBMR), this project investigates the complex pathological cascade of Arrhythmogenic Cardiomyopathy (ACM) to uncover new disease mechanisms and develop targeted therapies. By utilizing advanced in vitro, ex vivo, and in vivo models—including novel stem cell-derived 3D heart tissues—the research team focuses on three core areas:

  • Integration of Functional & Molecular Data: Mapping spatial and temporal gene expression alongside cardiac function to pinpoint the factors driving arrhythmias and fibrosis.

  • Cell-Type–Specific Mechanisms: Analyzing how specific cardiac cell subtypes and their interactions influence the manifestation and progression of the disease.

  • Therapeutic Strategies: Evaluating compounds designed to restore impaired cell-cell adhesion, which is a central driver of ACM pathogenesis.

To capture a comprehensive picture of ACM, the project utilizes a wide array of structural, functional, and omics-based techniques.

A critical component of this multimodal approach is the advanced molecular analysis. Heidi Tschanz-Lischer is collaborating closely on this project, bringing specialized expertise to support the transcriptomics investigations. Her work focuses specifically on spatial transcriptomics, as well as single-cell and single-nucleus RNA sequencing, to deeply profile the molecular landscape of the affected cardiac tissues.

Research: Ongoing Projects - Department for BioMedical Research (DBMR)

Automated Machine learning to identify drivers of microbial community composition

The compositions of microbial communities are of central importance for ecosystem functioning and host-associated health. Recent advancements in metagenomic sequencing have facilitated the acquisition of microbiome composition data. However, current analytical methods often fall short in exploring the environmental factors shaping these communities, especially in complex settings with numerous variables.

Given the crucial ecologic role of microbiomes and methodological gaps, this project aims to develop an analytic pipeline that explores environmental variables and ranks them by their association with the microbial community compositions at hand. The developed pipeline implements best practices of conditional data analysis (CoDa) and uses state-of-the-art ordination, clustering, and machine learning techniques to handle data with minimal supervision and assumptions.

Team: Ianis Vilela, Stephan Peischl
Co-Supervisors: Claudia Bank, Adamantia Kapopoulou

Local Adapation and Chromosomal Inversions

Nature showcases a fascinating mosaic of adaptations, with species evolving to adapt to their ever-changing environments. Chromosomal inversions, structural mutations that reverse gene order and link different gene variants, play a key role in this process. These inversions, often influenced by environmental factors like temperature and altitude, contribute to reproductive isolation and new species emergence. However, the exact mechanisms behind their contribution to local adaptation and the formation of new species remain unclear.

This project explores whether genomic structure facilitates the establishment of locally adaptive mutations or if natural selection and adaptive genome content drive the evolution of structural variation. By developing new statistical methods and mathematical models, and using machine learning to analyze data from sticklebacks and fruit flies, the project aims to demonstrate how chromosomal inversions contribute to local adaptation.

Team: Stefan Strütt, Karolina Wąchała, Janosch Imhof, Stephan Peischl
Project Partners: Thomas Flatt, Katie Peichel

Supported by SNSF grant 10001034 (SNSF Link)

Arx / OpenGenomeBrowser

The Problem: Today, sequencing of microbial genomes is cheap. However, extracting biological insights from this data is often hampered by a lack of tools that are scalable, interactive, and accessible to all researchers. Biologists need a way to easily manage, visualize, and query their data without deep computational expertise.

Our Approach: We developed OpenGenomeBrowser (now called Arx and developed by the startup Abrinca Genomics), a scalable, dataset-independent web platform for interactive genome data management and visualization. The platform is designed to bridge the gap between raw genomic data and actionable biological insights. It won the 2023 SIB Innovative Resource Award.

Thomas Roder

Autocycler Express

The Problem: Generating perfect, complete bacterial genome assemblies from long-read sequencing data is challenging. Automated assemblers often make errors like duplicating plasmids or failing to circularize sequences. While manual curation tools exist, they are typically labor-intensive and require command-line expertise, creating a bottleneck as sequencing output grows.

Our Approach: Autocycler Express is a browser-based user interface that streamlines and simplifies the manual curation of genome assemblies using the Autocycler pipeline. It allows for a rapid, iterative process of visualizing assembly graphs, identifying contigs (e.g., via BLAST), removing bad clusters, and cleaning the graph. This approach empowers biologists without programming skills to participate in the curation process, dramatically reducing manual effort while achieving high-quality, closed genomes ready for submission.

Thomas Roder

Scoary2: Scalable Microbial GWAS

The Problem: The functions of 40–60% of bacterial genes are unknown. Microbial genome-wide association studies (mGWAS) are a key method to link traits to genes and discover their function. However, existing tools were not designed to analyze the thousands of traits (e.g., from mass spectrometry) generated by high-throughput experiments.

Our Approach: To address this bottleneck, we developed Scoary2. It is a complete rewrite of its popular predecessor, designed for significantly improved performance, and, crucially, an interactive data exploration app. Thanks to these innovations, Scoary2 is the first tool that enables the robust and scalable study of large phenotypic datasets using mGWAS.