Tutorials and workshops

Overview

Spatial transcriptomics has emerged as a transformative technique in biology, revolutionizing our ability to study cellular organization and function within tissues. With varying spatial transcriptomics techniques, we can measure transcriptome-scale gene expression in a two-dimensionally resolved resolution. This tutorial will provide an introduction to the analysis of spatial transcriptomics data, equipping participants with the skills and knowledge to start handling, analysing, and interpreting spatial transcriptomics data sets. In the tutorial, there will be an overview of different spatial transcriptomics techniques, but the hands-on exercises will be using 10x Visium data.

Audience and requirements

The target audience are researchers that are interested to analyse spatial transcriptomics datasets. Attendees should have basic understanding of:

• The programming language R
• Dimensionality reduction techniques (PCA, UMAP)
• Clustering techniques

Dimensionality reduction and clustering techniques are taught in the SIB course ‘Single-Cell Transcriptomics with R’ and prerequisites can be obtained through self-learning at https://sib-swiss.github.io/single-cell-training/.

Organizers

• Heidi Tschanz-Lischer (Interfaculty Bioinformatics Unit, University of Bern)
• Geert van Geest (Bioinformatics Trainer at the Training group & computational biologist, Interfaculty Bioinformatics Unit, Bern)
• David Miguel Francisco Ferreira (Interfaculty Bioinformatics Unit, University of Bern)

Overview

Microbial communities can be found across a variety of Earth’s ecosystems. Correct identification and quantification of microbial community members, i.e. taxonomic profiling, is thus a key step to studying the role of microorganisms in ecosystem function and dysfunction. Decreasing costs have enabled the use of whole metagenomic sequencing approaches to study microbial communities, and the total number of publicly-available datasets has quadrupled over the last five years.

In this tutorial, we aim to raise awareness of key concepts, approaches, and statistical considerations for taxonomic profiling from whole metagenomic sequencing data. After tutorial completion, participants will be able to choose tools best suited to their research question and apply learnt concepts to answer a biological question on a dataset of thousands of taxonomic profiles.

Audience and requirements

We invite novice and intermediate users who regularly deal with whole metagenomic sequencing data. For novice users, the tutorial will cover best practices in the field and help avoid common pitfalls during taxonomic profiling of metagenomes. Intermediate users will have the opportunity to discuss their experiences during practical sessions and to work with thousands of pre-existing taxonomic profiles with associated metadata.

Attendees will be required to bring their own laptop (no tablets). They should be able to run tools on the command line and be familiar with working with data frames in Python and/or R.

Organizers

• Marija Dmitrijeva, Postdoctoral Researcher, Microbiome Research Lab, ETH Zurich
• Hans-Joachim Ruscheweyh, Senior Bioinformatician, Microbiome Research Lab, ETH Zurich

Overview

Due to the close relation between protein function and structure, obtaining the 3D structure of a protein of interest is a key step in biochemistry and molecular biology. Typically, this information is obtained using experimental or computational approaches, such as homology modelling or deep neural networks. Independently of the method used, the protein structures obtained are only models that best attempt to describe the data used to construct them, and thus are prone to errors. In this tutorial, we will revise the common and most important aspects of protein structure model quality assessment. We will provide practical tips on how to identify a problematic model, find a replacement, and run quality assessment and re-refinement tools to obtain a protein 3D structure adequate for your target downstream applications.

Schedule (tentative)

Audience and requirements

Any life science researcher interested in using protein structures in their work. No prior knowledge is required beyond basic knowledge of proteins.

Organizers

• Joana Pereira (Computational Structural Biology, Basel)
• Gerardo Tauriello (Team Lead, Software Development, Computational Structural Biology, Basel)

Overview

In this tutorial, we introduce the powerful and flexible TidyModels framework, a collection of packages for statistical modeling in R. We will show how to use TidyModels for machine learning applications with biological data, with emphasis on clean and understandable code. Participants will learn to perform data preprocessing, hyper-parameter tuning, model selection, and model interpretation using TidyModels, ensuring a comprehensive understanding of the entire data analysis pipeline. By the end of the workshop, participants will develop the skills to create reproducible machine learning models that can be seamlessly integrated into R data analysis workflows.  

Audience and requirements

This workshop is ideal for researchers who are looking to start their machine learning journey in R. Participants must be familiar with R programming. Previous knowledge of basic machine learning is recommended, but short introduction will be provided.

Organizer

• Ali Saadat (Host-Pathogen Genomics group, EPFL, Lausanne)
• Simon Tang (Host-Pathogen Genomics group, EPFL, Lausanne)

Overview

New spatial omics technologies are generating vast amounts of data, yet the analyses that utilize the spatial component are not straightforward. Most spatial omics approaches can be classified under high-throughput sequencing-based or imaging-based technologies. In terms of analysis, these two technological streams are very distinct. In this tutorial, we will explore the application of various spatial statistics tools to these two spatial omics streams. Pattern analysis for spatial omics data (PASTA) will highlight the usefulness and transferability of existing spatial statistics approaches in the context of spatial tissue profiling. Using a vignette that involves data from multiple technologies and an R package, concepts will be introduced, assumptions discussed and biological use cases will be shown with inline code.

Audience and requirements

The target audience for the tutorial are computational researchers that are faced with the challenge to analyse spatially-resolved omics data and interested in what spatial statistics has to offer.

Attendees must have a computer with R installed.

Organizers

• Martin Emons (PhD Student, Statistical Bioinformatics Group, University of Zurich)
  
• Samuel Gunz (PhD Student, Statistical Bioinformatics Group, University of Zurich)
• Mark D. Robinson, Professor of Statistical Genomics, UZH
• Helena L. Crowell, Postdoc, CNAG

Overview

Experimental and modelled 3D structures are widely used as the main source of information in the studies of the structure-activity relationships of proteins. However, post-translational modifications (PTM), including glycosylation, are often neglected even though they are known to play a major role in protein structure stability, solubility, protein-protein recognition, and resistance to aggression. This tutorial will demonstrate why and how crossing information regarding protein structures, sequences and glycosylation can help get a better understanding of protein structure and activity. It will also show the utility of such analyses in the field of personalized oncology and drug design. The tutorial will be given by the developers of different algorithms and databases developed by the SIB.

Tentative schedule

Audience and requirements

The workshop is intended for a wide audience including scientists ranging from master students to senior researchers.
Participants should bring their own laptop with a recent version of the Firefox web browser installed.

Organizers

• Fanny Krebs (Computer-aided Molecular Engineering group, Epalinges)
• Frédérique Lisacek (Group Leader, Proteome Informatics, Geneva)
• Ute Röhrig (Senior Research Scientist, Molecular Modelling group, Lausanne)
• Vincent Zoete (co-Group Leader, Molecular Modelling, Lausanne, and Group Leader, Computer-aided Molecular Engineering, Epalinges)

Overview

This workshop is aimed at presenting a set of software tools and good practices in reproducible research that are applied and supported by the Scientific IT Services of ETH Zurich. These tools and practices address the complete bioinformatic data life cycle of life science research: from initial experimental prototyping and production of -omics data in core facilities to the final publication of FAIR-compliant scientific results.

The experts of Scientific IT Services will share their extensive experience in the areas of data management, analysis and reproducibility with presentations and practical sessions, where participants will be able to actively work on example use-cases. The software toolbox presented includes managing and analysing data in openBIS with Electronic Lab Notebook, Jupyter, pyBIS Python module and Reproducible Research Platform (RRP) as well as Snakemake workflows development and publishing with various software stacks and computational environments.

Audience and requirements

The target audience includes potentially all members of SIB-affiliated labs.

Basic Python knowledge is beneficial for understanding and hands-on exercises in some modules.

Organizers

The workshop is organised by Scientific IT Services of ETH Zurich, which is an SIB-affiliated group led by Bernd Rinn.

• Caterina Barillari, PhD, member of the Research IT Platforms team of SIS. Manager of data management services provided by SIS to ETH research groups and to the Swiss academic community. Delivering regular data management workshops with the ETH Library and openBIS trainings.
• Rostyslav Kuziakiv, MD, PhD, member of the Research IT Platforms team of SIS. Since 2013 consulting, training, and supporting research groups on a wide array of research data management and data analysis concepts.
• Michal Okoniewski, PhD, member of the Computational Data and Science Support team of SIS. Since 2014 providing bioinformatics co-analysis support and teaching courses on high performance computing -omics applications and bioinformatic workflows for the ETH research community.

Overview

Sex and gender have historically been ignored as explanatory variables in most biomedical research. Reasons for this notably include the concern that inclusion of these variables might escalate experimental costs and complicate data analysis. However, in recent decades, several studies, ranging from research on cardiovascular diseases to immunity, showed that these are important explanatory variables which should not be ignored.

This workshop will focus on biological sex and discuss practical approaches for best incorporating it into the design and analysis of biomedical research experiments, with special emphasis on clinical and non-clinical animal experiments. We will first define what is biological sex in key model organisms, and highlight its multidimensional character, and its delineation from “gender” in the case of humans. We will then discuss experimental design and analytical approaches that can be used to evaluate differences between the sexes, or to effectively control for the effect of sex on other variables of interest.

Examples of concrete questions addressed in the workshop:

• How many male and female mice do we need for a given research project? What are the consequences of including more or fewer?
• Is a binary classification of samples into male and female categories always appropriate? When does it make sense to use other variables – such as hormonal levels or karyotypes – in place of, or in addition to, the biological sex variable?
• How to report the use (and limitation thereof) of sex as a variable in the method section of a paper and in a project proposal? Are there any standards for sex reporting in databases?

We aim at putting participants in a better position to understand the inherent sex biases embedded in experimental design choices and their impact in subsequent analyses; analyse datasets appropriately under the lens of the sex dimension and correct for sex biases where possible; carefully report the results and raise awareness among data collectors and colleagues about issues related to the inclusion of the sex dimension in biomedical research.

Audience and requirements

This tutorial is designed for bioinformaticians and computational biologists working in basic, translational and clinical research, with human and/or animal data, involved in planning/experimental design and/or performing downstream analysis.

Participants will need their personal laptops.

All the didactic material produced by BSC for this event (presentations, notebooks, etc.) will be shared with SIB organizers and participants, and it can be further distributed, adapted and reused in the future so long as attribution is given to the creator (CC-BY license).

Schedule (tentative)

Organizers

• Maïa Berman (Team Lead Communications, Lausanne, SIB Diversity Working group)
• Davide Cirillo (Life Sciences Department, Barcelona Supercomputing Center), external speaker
• Aitana Neves (Team Lead Data Science at SIB clinical bioinformatics, Lausanne, SIB Diversity Working group)
• Xavier Robin, (Senior Software Developer at Computational Structural Biology, Basel, SIB Diversity Working group)
• Julien Roux, (DBM Bioinformatics Core Facility, Basel, SIB Diversity Working group)
• Leonore Wigger (Senior Computational Biologist at Vital-IT, Lausanne, SIB Diversity Working group)

Confirmed instructors 

Davide Cirillo and Frédéric Schütz (Head of Biostatistics, Biostatistics platform)

Overview

Ever heard of the reproducibility crisis, the four horsemen of irreproducibility, or do you know what a pre-registered report is? The answers are irrelevant! Like climate change, nothing is more important than this…

We will venture together to a place many have spoken of, but few have gone to – trying to reproduce the results of scientific papers. After a quick introduction into reproducibility and the setup of the hackathon, the participants will be split into teams and assigned publications to reproduce. Teams will collaboratively find ways to reproduce, to gain access, and to discover new sources of tolerance, navigating through the (potential) hurdles of replicating papers. Each team will present their findings, detailing their journey of attempting to replicate the results in a final discussion round. This will include insights gained, obstacles encountered, and suggestions for improving reproducibility. We will be using the ReproHack Hub to select/submit papers and use their reporting functionality to share our reviews with the community.

This is not an introduction to programming or data analysis. Participants should have a favorite programming language and skills in analyzing data on their own. Participants should bring their own laptops (incl. power cords) and can suggest publications for reproduction (but do not have to).

Audience and requirements

This hackathon is ideal for PhD students, Postdocs, and research professionals who are eager to deepen their understanding of (computational) reproducibility in research and are committed to fostering a culture of integrity and excellence in science.

Attendees should be able to use R and/or Python, and have both installed on their laptop, and have some experience around statistics and computational data analysis. They should be comfortable with statistics, programming and a computational lingo used throughout the hackathon.

Organizers

The workshop will be organized by NEXUS Personalized Health Technologies and the Swiss Reproducibility Network. The Working Group for Computational Reproducibility together with members of the SwissRN Academy will be hosting and coaching the workshop.

For information and questions, please contact Daniel Stekhoven.

Overview

As the collection of multi-omics healthcare data grows, robust data analysis infrastructures become essential for sustainable knowledge extraction and for human health improvement. Apart from accessing data quality, consistency and reproducibility, this analysis infrastructure should provide tools for multi-omics data integration and model inference. The workshop aims to discuss best practices for single-cell gene expression analysis, focusing on single-cell RNA-seq. The first half will address scRNA-seq analysis requirements for multi-omic integration and biophysical process modeling. The second half will focus on analysis reproducibility and data/procedure sharing, aligning with FAIR principles and emphasizing cell type annotation reproducibility. Talks by experts will highlight current practices and challenges, followed by open discussions among participants. Ultimately, the workshop seeks to establish best practices and reproducibility principles for scRNA-seq analysis.

Audience and requirements

The workshop is targeted to experts and practitioners in the field of single-cell gene expression data, in particular scRNA-seq, including in particular bioinformaticians of core facilities who provide scRNA-seq data analysis on a daily basis.

Organizers

• Frederic Bastian – Associate Group Leader, University of Lausanne
• Fabrice David – Scientific collaborator in Bioinformatics, EPFL Lausanne
• Vincent Gardeux – Senior Scientist, EPFL Lausane
• Felix Naef – Group leader, professor, EPFL Lausanne
• Erik van Nimwegen – Group leader, professor, University of Basel
• Mihaela Zavolan – Group leader, professor, University of Basel

Overview

In an era where reducing carbon footprint is paramount across industries, research stands as no exception. The prevailing trend sees companies increasingly required to report on their energy usage and corresponding measurements. Our workshop offers a unique opportunity for participants to connect with peers, foster dialogues and exchange insights. Moreover, the workshop delves into programming techniques and methodologies aimed at optimizing energy efficiency. Attendees will depart equipped with practical tools to actively contribute to environmental sustainability within both our workplaces and the wider research landscape.

Audience and requirements

The workshop is addressed to bioinformatic scientists who are keen to understand our impact on the environment in their daily work and would like to learn and understand some fundamental aspects in coding to improve performance and to optimize usage of hardware. Also, it will be of interest to anyone who would like to share their views on how we can achieve more energy efficiency at work and aspects of environmental protection other than carbon footprint.

Organizers

• Qinyao Huang (bioinformatician, Bioinformatics Systems Biology group, University of Zurich)
• Samuel Neuenschwander (senior computational biologist at Vital-IT and Department of Computational Biology at University of Lausanne)
  
  

External speaker

• Rick Wertenbroek (engineer and PhD student at the University of Lausanne and School of Engineering and Management Vaud, Lausanne)