Tutorials & workshops

[FULLY BOOKED] Overview

beginner level –  single-cell transcriptomic – single-cell analysis ––  RNA Sequencing (scRNA-Veq)

Single cell RNA sequencing (scRNA-seq) has emerged as a revolutionary tool to measure the gene expression of complex biological systems at the level of individual cells. Its rapid development has facilitated the identification of previously unknown cell types in heterogeneous populations, empowering scientists to generate detailed tissue atlases describing the transcriptomic profiles of thousands or even millions of cells.

Although the single cell genomics field continues to rapidly develop, the lack of expertise on handling scRNA-seq datasets has restricted its use among the larger scientific community. This includes bioinformaticians who are only familiar with bulk sequencing data. The aim of this course is to enable researchers to start applying the fundamental scRNA-seq analysis pipeline to data generated in their labs. We will outline how to interpret results of a scRNA-seq dataset and, in teams, demonstrate the basics of data preprocessing and analysis using scanpy in Python. We will discuss common concerns in the field, including preprocessing choices, dimensionality reduction, cell type clustering and identification, batch effect correction, and pseudotime analysis. By the end of the tutorial, participants will be able to run scanpy analysis on their own datasets as well as understand how to overcome potential bottlenecks.

Learning objectives

The attendees of this tutorial will learn how to:

• Evaluate the quality of an scRNA-seq experiment
• Perform normalization, feature selection, dimensionality reduction and clustering with scanpy to identify cell types and differentially expressed genes
• Understand preprocessing choices typically made during single cell analysis
• Discuss and apply temporal methods such as pseudotime and RNA velocity

Schedule

Audience and requirements

Maximum number of participants: 25 

The target audience will be students and researchers who plan on using single cell RNA sequencing in their own projects, or who are interested in learning how to analyze such data but are unfamiliar with the approaches and pitfalls of single cell analysis. It is preferred if participants have some previous exposure to a programming language (such as Python), but no experience with single cell data analysis itself is required and attendees do not need to be frequent programmers.

Attendees will be required to bring their laptop to the tutorial as well as have completed the installation instructions for Python, Anaconda, and Jupyter notebooks that we will provide a couple weeks in advance of the tutorial data

Organizers

• Alex Lederer, PhD student in the laboratory of Neurodevelopmental Systems Biology with Dr. Gioele La Manno at EPFL, Switzerland
• Alexandre Coudray, PhD student in the Laboratory of Virology and Genetics with Prof. Didier Trono at EPFL, Switzerland

Overview

beginner level –  Spatial Transcriptomics  – Single-cell RNA sequencing

Single-cell RNA sequencing (scRNA-seq) technologies enable to define the gene expression profiles of single cells allowing the characterisation of heterogeneous cell populations. However, scRNA-seqlacks the spatial context as the tissue must be dissociated. To overcome this issue, spatial transcriptomics (ST) is a new and evolving technology that measures transcriptomic information while preserving spatial information. Consequently, ST can unravel the complexity of biology, especially the molecular mechanisms behind diseases that cannot be understood by other technologies.

There are different ST techniques with different technical parameters, and there is often a trade-off between the number of genes profiled and the efficiency of the technique. In this tutorial, we will present an overview of the different currently available methods to perform ST, their strengths and drawbacks. The focus will be on the two main methods: spot-based (10x Visium) and in-situ (10x Xenium) technologies. Spot-based technology allows the detection of the whole transcriptome across the entire slice of tissue with high sensitivity and specificity, but it's limited to capture spot resolution. In contrast, the newest in-situ technology Xenium released at the end of 2022 allows us to have an extremely high resolution at subcellular level, but it can be biased due to the requirement of pre-selected gene targets. These two technologies can be complementary as Xenium data can be used to interpolate the cell composition of Visium spots and leverage Visium whole transcriptomics to understand tissue heterogeneity and discover new biomarkers.

Therefore, in this tutorial, we will use computational tools to analyse single-cell RNA-seq, Visium and Xenium data to demonstrate how whole transcriptome and targeted in-situ data can be integrated to provide highly complementary and additive biological information. The right balance between lectures and practical training will provide entry-level guidance for participants interested in ST data analysis.

Learning objectives

At the end of this tutorial, attendees will understand how the different ST technologies work, and the advantages and limitations of each of the methods. This tutorial will provide the participants with theory and practical training to analyse ST data (Visium and Xenium), as well as a guide to the most recent computational tools to perform more advanced ST analysis. With the provided guidance of the instructors, participants will be able to independently analyse different ST data.

Schedule

Audience and requirements

Maximum number of participants: 30

This tutorial is designed for individuals interested in spatial transcriptomics data analysis. Participants must possess a basic understanding of R and RNA-seq gene expression method and data format. Participants should bring a laptop with Wi-Fi, the latest versions of R, Rstudio, and relevant R packages installed. Details on package installation will be sent via email before the tutorial.

Organizers

• Ana Cristina Guerra de Souza, PhD, Bioinformatician at the Translational Data Science Facility (SIB), Switzerland
• Tania Wyss, PhD, Bioinformatician at Dept of Fundamental Oncology (UNIL) and at the Translational Data Science Facility (SIB), Switzerland

Overview

beginner level –  Spatial Transcriptomics  –  Single-cell RNA sequencing

There are several reasons why learning or improving SQL (Structured Query Language) can be beneficial for data science:
1. SQL is the standard language for working with relational databases, which are a common source of data for data science projects. By learning SQL, you'll be able to easily retrieve, manipulate, and analyze the data stored in these databases.
2. SQL allows you to efficiently work with large datasets. When working with data science projects, it is common to deal with datasets that are too large to fit into memory. SQL provides powerful tools for filtering, grouping, and aggregating large datasets, so you can work with subsets of the data that are small enough to fit into memory.
3. SQL can help you to understand the underlying data better. Even though SQL is a programming language, it is declarative, unlike most of the general-purpose programming languages like Python, C++ or Java. It means the code you write will tell the SQL engine what you want and the engine will take care of the HOW to get the results, so it will help to get more understanding about the dataset, its structure and its constraints
4. SQL is a good tool for data preparation, cleaning and validation. Since it is a powerful tool to manipulate and filter the data, you could use it to prepare your dataset to a better shape before applying any statistical or machine learning models.
5. SQL can be a valuable skill in the job market. Many companies store data in relational databases, and SQL knowledge is often a required or preferred skill for data science positions.

In summary, learning SQL can be a valuable addition to your data science toolkit, as it allows you to efficiently work with large datasets and can be a valuable skill in the job market. Take a look at the results of this survey.

Schedule

Audience and requirements

Maximum number of participants: 20

This tutorial is designed for data scientists, programmers, bioinformaticians or students. Participants should bring a laptop and install the DBMS and download the database prior to the tutorial (still under discussion).

Organizers

• Dillenn Terumalai, Software developer at SIB clinical bioinformatics, Switzerland
• Florent Tassy, Data manager and software developer at SIB clinical bioinformatics, Switzerland
  

Overview

Intermediary level –  FAIR principles  

The FAIR (Findable, Accessible, Interoperable and Reusable) principles provide guidelines for making research data and other resources more easily discoverable and reusable, which can help increase your research's impact and exposure. Adhering to these principles also ensures that your research is more reliable and reproducible, as others can more easily access and provide feedback. In addition, making your research FAIR can promote the principles of open science and make it easier for others to contribute to and build upon your work. Finally, many funding agencies and journals now expect that your research outputs be made FAIR as a condition of funding or publication, so adhering to these principles can help to ensure that your research meets these requirements.

Sharing and reusing data, software, and documentation is essential to the FAIR principles and should be a routine task for scientists. Researchers can achieve this by providing all necessary information (data, software and parameters used, scripts for the analysis, databases and their versions) and using markdown to create a single file that can be easily shared as a web page.

Learning objectives

In this tutorial, participants will learn tools and concepts to take significant steps towards adhering to the FAIR principles, which will enhance the benefits of sharing and collaboration in research. More specifically, they will learn:

• Create notebooks and websites based on Markdown, and Python or R with Quarto
• Use Git and GitHub to version control the generated content
• Host a website by making use of GitHub actions and GitHub pages
•  Link the GitHub repository to Zenodo and give it a unique identifier (DOI)

Schedule

Audience and requirements

Maximum number of participants: 30

This tutorial is aimed at computational biologists, bioinformaticians, researchers, scientists and trainers working in the life sciences who want to learn how to make their research and training FAIRer with reproducible notebooks and websites. Participants are expected to have an introductory level in programming with R or Python. Participants should have a GitHub account and bring their laptops with either the latest versions of RStudio or VSCode pre-installed.

Organizers

• Geert van Geest, Bioinformatics Trainer & Support Specialist (SIB), Bioinformatician and data analyst at Interfaculty Bioinformatics Unit of the University of Bern, Switzerland
• Wandrille Duchemin, Bioinformatics Trainer & Computational Biologist (SIB), Bioinformatician and data analyst at the sciCORE core facility of the University of Basel, Switzerland

Overview

Intermediate level ––  Method benchmarking  ––  benchmarks  ––  omics datasets

The rise of large-scale omics datasets has led to a growing number of methods to model and interpret them, making it hard to identify performant computational methods to use in discovery research. Method benchmarking is critical to dissect important steps of an analysis pipeline, formally assess performance across common situations and edge cases, and ultimately guide users. While some form of comparison is standard practice in method development, current approaches have several limitations (Sonrel et al. 2022). One main issue is that benchmarks are not easily extensible and with the constant emergence of new approaches, this often leads to rapidly outdated or even contradicting and irreproducible conclusions. We believe that if benchmarks were organized in a systematic way and conducted at a higher standard, they will have a considerably greater impact in guiding best practice in employing computational methods for discovery research.

To address these issues, we created Omnibenchmark, a modular and extensible framework based on the free open-source platform renku. The framework connects data, method, and metric modules via a knowledge graph, which tracks relevant metadata (e.g. software environment, parameters and commands). Results can be viewed in a dashboard or openly accessed, and new modules can be added easily with pre-configured templates. To facilitate community contributions for adding new reference datasets, evaluation metrics or computational methods, we maintain a series of pre-configured templates. Each element of Omnibenchmark is packaged with all dependencies and can be inspected, re-used, modified, and integrated onto other platforms in compliance with the FAIR principles.

Learning objectives

At the end of the tutorial, participants should be able to:

• understand components of the Omnibenchmark framework
• submit new reference datasets, methods or metrics modules
• visualize and re-use output results and performance summaries

Schedule

Audience and requirements

Maximum number of participants: 20

This tutorial is addressed to computational-minded researchers who are tasked with benchmarking methods that they are familiar with (or that they developed) as well as for benchmarkers that would like to port their work to our system.

Participants are expected to have :
- Intermediate knowledge of R or Python
- Basic knowledge of UNIX and version control systems (Git)
- Personal laptop with Wifi connection
- Code for method(s) and/ or benchmark(s)- (optional)

Organizers

• Main presenters: Almut Lütge and Anthony Sonrel, PhD Students and co-developers of Omnibenchmark, Statistical Bioinformatics Group at the University of Zurich, Switzerland
• Technical collaborators and knowledge exchange: Dr. Izaskun Mallona, Statistical Bioinformatics Group at the University of Zurich (co-developer of Omnibenchmark); Dr. Charlotte Soneson, Research Associate, Computational Biology Platform, FMI and SIB Swiss Institute of Bioinformatics, Switzerland
• Introductions and support: Mark D. Robinson, Associate Professor, University of Zurich and SIB Group Leader, Switzerland

Overview

Beginner level –  genome annotations – orthologs

The Ensembl project provides freely available access to genome annotation datasets including gene, variant and regulatory feature annotation as well as comparative genomics analyses for over 300 vertebrate species and 30'000 non-vertebrate eukaryotes and prokaryotes. All of the data can be retrieved through Ensembl’s online genome browsers (www.ensembl.org, www.ensemblgenomes.org and rapid.ensembl.org) as well as programmatically via Ensembl’s REST API.

This tutorial will introduce you to the range of data available through Ensembl and the concepts of the Ensembl REST API and guide you through the principles of retrieving Ensembl data programmatically using both Python and R.

To participate in the hands-on aspects of this tutorial, including writing and executing REST API queries, you will need to bring a laptop. Tutorial materials, including slides, screenshots, exercises, sample files and solutions will be available before the tutorial and will remain permanently online at the Ensembl training portal: https://training.ensembl.org.

Learning objectives

● Outline the different data types available through Ensembl.
● Identify the appropriate methods for data retrieval from Ensembl.
● Perform queries and extract data returned from the Ensembl REST API.

Schedule

Target audience and requirements

Maximum number of participants: 30

The tutorial is aimed at new and existing Ensembl users, from both the wet-lab and bioinformatics communities. The tutorial is designed to provide participants with a greater understanding of the data available through the Ensembl interfaces and how to efficiently retrieve it at various scales. There are no prerequisites for this tutorial, although a basic understanding of programming with Python or R would be beneficial. For the interactive aspects of this tutorial, participants are required to bring their personal laptops. Tutorial materials, including slides, screenshots, exercises, sample files and solutions will be available before the tutorial and will remain permanently online at the Ensembl training portal.

Organizers

Louisse Mirabueno, European Molecular Biology Laboratories - European Bioinformatics Institute (EMBL-EBI), UK

Louisse Mirabueno completed a Bachelor’s degree in Genetics and Cell Biology at Dublin City University, and obtained an MPhil. at the University of Reading. She has worked in both academic institutes and industry. Her experience ranges from genomic analyses in vertebrates and bacterial populations, high-throughput screening in biotech and early-stage drug discovery, and research and development in long-read sequencing. Louisse joined the Ensembl Outreach team in January 2022 and has since delivered 28 in-person and virtual workshops teaching participants how they can integrate Ensembl data in their research.

Overview

Beginner level –  Long-read sequencing – genome annotations – orthologs

Today, bacterial genomes can be sequenced for less than 200 dollars per genome. If sequenced with long-read technologies, it is often possible to assemble the data into complete bacterial chromosomes including plasmids. In this beginner-friendly tutorial, we will show and explain the steps required to go from raw sequence data to an assembled and annotated genome made accessible through OpenGenomeBrowser. To provide a hands-on experience, each participant will receive a PacBio HiFi raw dataset of one bacterial strain, with which the entire workflow will be demonstrated. In the end, the products of all participants will be collected, distributed to everyone, and imported into OpenGenomeBrowser. 

Because the software used in this tutorial are resource-efficient algorithms and are provided as Docker containers, participants will be able to follow the tutorial on their own laptops or using a shell in the cloud. In case one step in the pipeline does not work, the intermediate results will be made available so that the participant can continue the tutorial. 

The following tools will be used during the tutorial:
- Assembly: La Jolla Assembler (LJA) (Bankevich et al. 2022)
- Polishing: HyPo (Kundu et al. 2019)
- Structural annotations: bakta (Ekim et al. 2021)
Import into genome database: OpenGenomeBrowser (Roder et al. 2022)

Learning objectives

The objectives of the tutorial are:

• Getting an overview of a microbial genome assembly and annotation pipeline
• Understanding challenges in microbial genome data management
• Understanding orthologs
• Gaining experience using Docker: basic data processing and hosting the OpenGenomeBrowser stack (3 Docker images coordinated using Docker Compose)

Schedule

Target audience and requirements

Maximum number of participants: 30

The tutorial is aimed at computational biologists, bioinformaticians, researchers and trainers working in life science who want to learn about the basics of microbial genomics. Participants are expected to have basic experience in Bash. If they want to work locally, participants should install the latest version of docker beforehand. If not, any laptop can be used to follow the tutorial in the cloud. 

Organizers

• Thomas Roder, PhD student at analyst at Interfaculty Bioinformatics Unit of the University of Bern, Switzerland
• Marco Kreuzer, Bioinformatician and data analyst at Interfaculty Bioinformatics Unit of the University of Bern, Switzerland
• Heidi Tschanz-Lischer, Bioinformatician and data analyst at Interfaculty Bioinformatics Unit of the University of Bern, Switzerland

Overview

Beginner level –  systems biology – modelling –  digital twins  – data annotation and curation  – personalized medicine 

Digital twins are an emergent concept in Computational Systems Biology and personalized medicine. Many initiatives manifest an interest of creating computational models of increasing complexity that could be used to represent virtual patients and help in the decision-making regarding the appropriate treatment. Having robust and reliable computational models spanning all biological layers such as gene expression, signalling and metabolism to name a few, could revolutionize the way we treat Big Data for the benefit of precision medicine and improved medical care - tailored to the needs of each patient. As the number of computational models rapidly increases, the production of data is ever growing and the approaches, both mechanistic and AI-based, are rapidly developing, discussions about challenges and best practices are needed more than ever.

The purpose of the workshop is to bring together researchers working in the field of digital twins in computational systems biology who use various formalisms to address challenges of data integration and model personalization. The focus will be on presenting the state of the art in the field and how mechanistic computational modelling can upscale benefiting from AI-based methodological advances.

Several experts on computational modelling for personalized medicine, AI-based data integration and clinical digital twins will be invited (see list of tentative speakers below). In addition, a call for abstract submissions will be issued.

List of tentative invited speakers:

• Liesbet Geris, VPH Institute, BE
• Maria Rodriguez Martinez, IBM Zurich, CH
• Mariano Vazquez, BSC, ES
  
• Kristin Reiche, Fraunhofer Institute for Cell Therapy and Immunology, Leipzig, DE
• Jurgen Kuball or Zsolt Sebestyen, UMC Utrecht Cancer Centre, NL
• Ioannis Xenarios, UNIL/CHUV/Health2030 Genome Center, CH

Schedule

The morning sessions will be dedicated to AI-based approaches for medical digital twins; the afternoon sessions will cover the challenges of the field, addressing both ethical and technical aspects (storage, access and protection of sensitive data, calculation power computational frameworks). Each session will feature one invited talk from a top expert in the field.

Target audience and requirements

Maximum number of participants: 40

This workshop addresses students and researchers interested in learning about this emerging and promising concept of digital twins, giving them a unique opportunity to discover and discuss the state of the art in the field.

Call for abstracts

The organizers would like to invite interested scientists to submit an abstract for a selected talk. The abstract should be in pdf form, 400 words maximum with figures if needed, and not exceeding 1 page. Submissions should be sent to laurence.calzone@curie.fr with title: [BC]2 workshop abstract submission

Deadlines

• Abstracts sent by: 20 June 2023
• Decision on selected abstracts: 28 June 2023
• [BC]² workshop early-bird registration deadline: 30 June 2023
• Workshop: 11 September 2023

Organizers

• Anna Niarakis (Associate Professor; UEVE, Univ Paris-Saclay & INRIA Saclay; France)
• Laurence Calzone (Research Scientist; Institut Curie, U900 INSERM & Mines Paris Tech; France)
• Arnau Montagud (Postdoctoral Researcher at the Barcelona Supercomputing Center; Spain)

Overview

Single-cell functional genomics is a rising field of biology, which is bringing major insight into the life sciences. Single-cell data are rapidly increasing both in quantity and in diversity, but lack method and metadata standardization. While some large projects have clear standards of reporting, most datasets in biological databases have partial or non-standardized metadata. This leads to multiple non-compatible standards across datasets, and limits reusability, which in turn presents challenges to making these data useful to an increasing community of specialists and non-specialists.

There is a need for standards in the way single-cell data are stored and annotated, especially for cell type and other associated information. Indeed, metadata is critical to the capacity to use these large and potentially very informative datasets. It includes protocols, which constrain which transcripts were accessible or which normalizations are relevant, the association between barcodes and annotations, or the methods used to identify cell types. Existing ontologies and controlled vocabularies are not used systematically, even when information is reported.

To address these limitations, the funded project scFAIR has the mission to build a collaborative platform supporting and disseminating ORD practices for the single-cell genomics community, both for sharing datasets and their metadata, and for standardizing the way data are shared across datasets. During this workshop, participants will have the opportunity to shape the definition of this collaborative platform, by sharing their needs and best practices. This workshop will notably cover:

• how to describe and capture information about processing pipelines, in order to reproduce analyses, and determine how cell features were identified;
• how to provide in a standard way information about barcodes identifying cells, and the raw data of the analysis;
•  how to annotate cell clusters with controlled vocabularies and ontologies;
•  best practices in quality control;
•  best practices in cell type inference.

 
On each topic, after an introduction by a field expert, we will ask participants to share the difficulties they face in their own work, and what they miss to leverage to their full potential existing single-cell data. We will then work together to synthesize recommendations. We expect that, as a result, participants will acquire an improved knowledge of the existing landscape in single-cell biocuration, and a collaborative involvement in the future of Open Research Data in scFAIR.

Contributions 

Selection mode: both invitation and call for submissions
The organizers will also contribute presentations. We will also invite experts in the single-cell field, from the joint working group between SIB and AGORA Cancer Research Center.

Invited speakers

• Jason Hilton, Stanford University
  Jason is the project lead for Lattice, a data coordination project supporting the Human Cell Atlas, and part of the team managing the CZI CELLxGENE resource.
• David Osumi-Sutherland, EMBL-EBI
  David leads a group of Ontology editors and developers within the Samples, Phenotypes and Ontologies (SPOT) team at the EBI, providing semantic solutions for annotating and organizing biomedical data following FAIR principles. His team notably contributes to the Cell Ontology and Provisional Cell Ontology.

Target audience and requirements

Maximum number of participants: 20

The workshop targets biologists, bioinformaticians, or curators, working or intending to work on single-cell data. 

Organizers

• Frédéric Bastian: associate group leader of the SIB team developing the Bgee database, specialized in gene expression representation, biocuration and bio-ontologies.
• Vincent Gardeux: Senior scientist at Deplancke’s lab at EPFL, developing the ASAP portal for single-cell RNA-seq analysis.
• Marc Robinson-Rechavi: professor of bioinformatics at the department of ecology and evolution of UNIL, and group leader of the SIB team developing the Bgee database.