The Oslo Bioinformatics Workshop Week 2024 is supported by UiO:Life Science, the Group for temporary employees at the Division of Laboratory Medicine (KLM TempAware), NCMM - Norsk senter for molekylærmedisin, Nordic Computational Biology and ELIXIR Norway.
Agenda
Building Customized, Interactive Genome Visualizations with GenomeSpy
Date: Wednesday 11 December 2024 9:00-12:00 Add to calendar
Room: Python (room 2269)
Description
Most genome visualization tools are designed for specific data formats and offer limited customization options, often leaving researchers to build entirely new solutions from scratch for tailored visualizations. GenomeSpy (https://genomespy.app/) is a toolkit that addresses this gap by providing a flexible way to create interactive genome visualizations. Built on a visualization grammar, its declarative language allows users to define custom visualizations without extensive programming. Additionally, its unique GPU-accelerated architecture ensures smooth, responsive interactions, even with large datasets. In this workshop, we will explore the types of visualization challenges that GenomeSpy can address and introduce the fundamentals of its visualization grammar. In the hands-on tutorial, participants will start by exploring a high-grade serous carcinoma dataset using an existing GenomeSpy visualization, demonstrating its capacity to analyze large sample collections effectively. Following this, we will implement simple visualizations using the GenomeSpy Playground application and ObservableHQ notebooks. Additionally, we’ll set up a local website for GenomeSpy, supporting larger-scale visualization projects that can later be deployed on a private or public web server.
Learning outcomes
By attending this workshop, participants will:- Understand the concept of visualization grammars
- Gain insight into the capabilities of GenomeSpy for genomic data
- Learn to set up GenomeSpy for customized visualizations
- Know how to specify and create basic genome visualizations using GenomeSpy
Target audience
This workshop is suitable for anyone interested in data visualizations, especially those exploring large genomic datasets, bioinformaticians, and computational biologists.
Pre-requisites
- Familiarity with JSON (JavaScript Object Notation). https://www.tutorialspoint.com/json/index.htm
- Prior knowledge of visualization grammars such as ggplot2 or Vega-Lite is helpful but not required.
Equipment to bring
A laptop with a code editor that supports JSON Schema, such as Visual Studio Code. It should also have python or nodejs installed to launch a local web server.
Instructor(s):
Kari Lavikka (main)
Predicting drug combinations using only single-cell transcriptomics
Date: Thursday 12 December 2024 9:00-11:00 Add to calendar
Room: Python (room 2269)
Description
The challenge of cancer treatment lies in the vast number of potential drug combinations and dosing options. Each cancer type presents unique biological characteristics, and tumors often contain diverse cell populations that respond differently to therapies. This complexity leads to a computational explosion in identifying the most effective treatments. Advanced computational tools are essential for navigating this landscape. Here, we introduce scTherapy, a new machine learning tool that addresses these challenges by prioritizing multi-targeting therapies for individual cancer patients. It uses single-cell transcriptomic profiles to identify patient-specific treatment options that effectively co-inhibit cancer cells while minimizing harm to normal cells. Experimental validations show that 96% of the predicted multi-targeting treatments demonstrate selective efficacy or synergy, with 83% exhibiting low toxicity. Participants will learn about scTherapy's approach to analyzing single-cell RNA sequencing (scRNA-seq) data and utilizing phenotypic response profiles to identify potential therapies that target specific tumor subclones, taking into account the heterogeneity within and across patient samples. Through hands-on application, attendees will understand how to leverage the pre-trained machine learning model within scTherapy to predict compound and dose pairs, ensuring effective and low-toxicity treatment options for individual patients with solid or hematological malignancies.
Learning outcomes
- After attending this workshop, learners will gain a comprehensive understanding of how to leverage single-cell RNA sequencing (scRNA-seq) data for precision oncology and personalized treatment design. Participants will learn to use scTherapy and its modules—encompassing scRNA-seq preprocessing, cell-type annotation, subclone identification, drug-dose prediction, —to analyze patient-specific tumor profiles and predict targeted, effective therapies. Through this workshop, they will learn to interpret model outputs and identify drug candidates with minimal toxicity and maximal efficacy.
Target audience
This workshop may be particularly beneficial for researchers, graduate students, and professionals interested in single-cell transcriptomics, machine learning applications in cancer therapy, and personalized medicine approaches.
Pre-requisites
- Experience in single-cell transcriptomic analysis, especially with tools like Seurat
- Basics in R programming
- Access to R (version 4.0 or above) and Docker desktop
Equipment to bring
A laptop with docker desktop downloaded
Instructor(s):
Kristen Nader (main)
AlphaFold and cryoEM for biomolecular structures
Date: Friday 13 December 2024 9:00-12:00 Add to calendar
Room: Prolog (room 2465)
Description
In this workshop you will learn how to use AlphaFold3 to predict the structure of proteins. The workshop provides examples of protein-protein complexes, proteins with PTMs, protein-DNA complexes and proteins in complex with simple ligands. You will learn how to interpret AlphaFold results in terms of confidence. If you have a Fox account, you will also learn how to use the AlphaFold2 package installed on the Fox server. We will also briefly explain how to access and use the software for processing experimental cryoEM data using the Fox server.
Learning outcomes
After attending this workshop, learners are able to:- run and interpret results of web based AlphaFold3
- set-up and run AlphaFold2 installed on Fox server
- visualize, analyze, and compare predicted models using ChimeraX
- have knowledge on how to find and run Relion for cryoEM processing on Fox server
Target audience
The workshop is suitable for all career stages.
Pre-requisites
- Programming experience is not required, but participants should be familiar with using the terminal to run Fox-based programs. The first part of the workshop will focus on the web-based AlphaFold, which does not require a Fox account and use of terminal. All participants should familiarize themselves with ChimeraX.
Equipment to bring
Participants should bring their own laptop with installed ChimeraX for protein visualization. A uio account is also required to access Fox.
Instructor(s):
Nikolina Sekulic (main)
Ahmad Ali-Ahmad
Accelerating Genomics Workflows on GPUs: Hands-on with NVIDIA Parabricks
Date: Thursday 12 December 2024 9:00-11:00 Add to calendar
Room: Prolog (room 2465)
Description
In the fast-evolving world of genomics, Next-Generation Sequencing (NGS) analysis pipelines are pivotal for the comprehensive evaluation of patient genetic makeup in clinical and research settings. Traditional bioinformatic tools optimized for CPU-based computing often result in prolonged processing times that can extend to several days, creating significant bottlenecks as population genomics studies and clinical practices scale. To address these challenges, accelerated NGS pipelines leveraging graphical processing units (GPUs) have emerged as transformative solutions. This two-hour workshop aims to demonstrate the capabilities of NVIDIA Parabricks tools in NGS analysis. The session will commence with a concise overview of the NVIDIA Parabricks software suite, a leading toolset known for its remarkable ability to drastically reduce computational time while maintaining exceptional precision and recall in variant calling. Attendees will gain insights into the workflow of the Parabricks germline variant calling pipeline. Following the overview, participants will engage in a hands-on session designed to provide practical experience with Parabricks. The workshop will conclude with a brief discussion addressing the advantages and potential limitations of Parabricks. This workshop is ideal for bioinformaticians seeking to enhance their expertise and efficiency in NGS analysis through cutting-edge technological advancements.
Learning outcomes
- Gain a foundational understanding of the NVIDIA Parabricks software suite and its application in germline variant calling.
- Acquire practical experience in using Parabricks tools for NGS analysis through a hands-on session.
Target audience
Bioinformaticians with above prerequisites
Pre-requisites
- Familiarity with NGS Analysis Workflows: A basic understanding of the next-generation sequencing (NGS) analysis process and its various stages, such as read alignment, variant calling, and result interpretation.
- Ability to Run Commands in a Linux Environment: Proficiency in navigating and executing commands in a Linux-based operating system, given that many bioinformatics tools and pipelines operate within this environment.
- Experience with Container Technologies: Knowledge of containerization tools such as Docker or Singularity, which are commonly used to encapsulate and deploy bioinformatics software.
- Basic Understanding of Genomic Data Formats: Familiarity with standard genomic data formats such as FASTQ, BAM, and VCF, which are integral to the NGS analysis process.
- Prior Experience with Bioinformatics Software: Some hands-on experience with bioinformatics tools and software, even if only at a basic level, will be beneficial.
Equipment to bring
MAC or laptop with linux OS (Windows laptop that can run SSH command and connect to a remote server. However we may not be able to provide support for Windows users)
Instructor(s):
Pubudu Samarakoon (main)
Workflow management and sharing with Nextflow and nf-core
Date: Tuesday 10 December 2024 9:00-12:00 Add to calendar
Room: Prolog (room 2465)
Description
Managing bioinformatics workflows (e.g. sequencing analysis pipeline) is a central problem for many of us. Simple bash scripts are maybe easy to write, but hard to run and maintain. Snakemake is a good solution for reproducibility and automation, but it is hard to reuse its components. Nextflow is a workflow manager tool like Snakemake, but solves those problems too that make Snakemake sometimes difficult to run. The price to pay is that getting into Nextflow might be hard. Here, we would like to reduce the initial time investment and help you understand the main concepts while you try building a small pipeline.
Learning outcomes
- After attending the workshop you will be more confident in starting your next project in Nextflow, and keep learning more about its ways of solving various challenges. We will explain the logic and basics of Nextflow. We will guide you for browsing among the available collection of stable pipelines and modules provided by nf-core. Through an example project of processing paired-end sequencing reads for CAGE analysis, you will have hands-on experience about building a pipeline by lego-ing together modules published in nf-core and adding your own.
Target audience
Bioinformaticians of any stage, field or interest are welcome. This is a technical workshop which can benefit anyone.
Pre-requisites
- It will be easier for you to appreciate the solutions of Nextflow if you have experience with pipelines, so that you experienced the common challenges. We will also make some comparisons to Snakemake. Neither of these are strong requirements as we will explain the key points through an example.
Equipment to bring
Attendees should bring their own laptop.
Instructor(s):
Katalin Ferenc (main)
Ieva Rauluseviciute
Using the SiSaNA Pipeline to Reconstruct and Analyze Patient-Specific Gene Regulatory Networks
Date: Monday 09 December 2024 13:00-15:00 Add to calendar
Room: Prolog (room 2465)
Description
As biologists, we frequently have a lot of data to work with but need user-friendly tools to understand the secrets our data holds. LIONESS (Linear Interpolation to Obtain Network Estimates for Single Samples; Kuijjer et al., 2019) provided a novel method for modeling gene regulatory networks, which model the interactions between transcription factors and target genes. A new tool called Single Sample Network Analysis (SiSaNA) has recently been developed, which uses the LIONESS methodology to create networks, but allows the user to both create and analyze such networks using only the command line, omitting any need for programming expertise. This workshop will first teach participants how to use transcriptomic data to reconstruct single-sample gene regulatory networks using SiSaNA. Then, participants will learn how to use SiSaNA to understand the information presented in these networks, as well as how to compare features of these networks between sample groups to identify key genes, transcription factors, and pathways of interest that distinguish these groups from one another.
Learning outcomes
- After attending this workshop, learners will know how to use SiSaNA to create and analyze single-sample regulatory networks. They will have experience running the example commands on the data provided in SiSaNA and will understand what a gene regulatory network is, as well as be able to produce publication-ready plots using the SiSaNA command line interface tool.
Target audience
Individuals of any career stage are encouraged to join. The LIONESS method used in SiSaNA is developed for use of transcriptomic data (microarray, RNASeq) so anyone who has that type of data and is interested in understanding the gene regulation within their samples should consider joining.
Pre-requisites
- For this workshop, users should have access to a UNIX terminal and UNIX environment. Users should be familiar with how to navigate on a UNIX terminal as well.
Equipment to bring
A laptop is highly recommended. While I will perform each step on screen as well, it would be great if participants could follow along on their own laptops as well.
Instructor(s):
Nolan Newman (main)
From data to publication: how ELIXIR Norway can help you for free
Date: Monday 09 December 2024 9:00-12:00 Add to calendar
Room: Prolog (room 2465)
Description
Are you working on a PhD or PostDoc project or applying for funding? Most projects now require a Data Management Plan (DMP), and ELIXIR Norway offers support—often at no cost! Life science research should also follow the FAIR Data principles (findability, accessibility, interoperability, and reusability), but implementing these can be challenging. ELIXIR Norway can assist with this and provide free data storage and analysis pipelines for non-sensitive data. This workshop offers hands-on training for creating DMPs using ELIXIR’s Data Stewardship Wizard (DSW). The workshop also introduces NeLS and usegalaxy.no, two free ELIXIR Norway services for data storage, analysis, and reproducible workflows.
Learning outcomes
- Understand what ELIXIR Europe & ELIXIR Norway is, and how these institutions can help you in your research
- Understand the relevant information to be captured in a DMP Start a new DMP on DSW, from scratch or from a set of pre-filled recommendations
- Share a DMP and work collaboratively on it
- Assess good practices in research data management and FAIRness
- Export a DMP according to formal requirements from funding bodies such as the Research Council of Norway, Science Europe, and Horizon Europe
- Transfer files from local machines to NeLS
- Transfer files between UseGalaxy and NeLS
- Understand how to use galaxy tools for data analysis
- Create galaxy "histories" for reproducible analysis
- Create, run and share workflows on UseGalaxy
Target audience
This workshop targets researchers at any career stage (PhD candidates, Post Doctoral Fellows, Researchers, and Professors) and especially personnel responsible for the data generated within a project. Technical personnel are also welcome.
Pre-requisites
- A user account on the Norwegian instance of DSW (https://norway.dsw.elixir-europe.org/)
- Feide login credentials are required to log in. Participants who do not have access to such credentials should apply in advance for a NeLS identity (see nels.bioinfo.no)
Equipment to bring
Laptop.
Instructor(s):
Federico Bianchini (main)
Jon Lærdahl
Viral identification, quantification and functional annotation using shotgun metagenomic sequencing data analysis
Date: Wednesday 11 December 2024 9:00-12:00 Add to calendar
Room: Prolog (room 2465)
Description
In this workshop we will give an overview of the existing tools for viral identification and functional annotation using shotgun metagenomic sequencing data. We will also present VirMake, a pipeline automating virome data analysis developed in the Rounge group. We will use a mock viral community to demonstrate the use of VirMake pipeline and further have a hands on tutorial on the downstream analysis of the output on viral taxonomy, quantification, and functions. Here the audience will learn the basics of the virome community analyses, including diversity and abundance estimation and normalization, beta diversity (i.e. diversity between samples) estimation, in addition to prophage characterisation, potential host identification, and functional characterisation including auxiliary metabolic genes. We will also discuss current limitations with viral data analysis.
Learning outcomes
- After attending this workshop, learners will have an overview of the existing tools and pipelines for virus identification and annotation, and have hands-on experience with analysis of viral data using shotgun metagenome sequencing.
Target audience
MSc students, PhD students and researchers interested in viral communities/virome data analysis
Pre-requisites
- Rstudio, knowledge of Snakemake/python can be beneficial but not required
Equipment to bring
A laptop with preinstalled R/Rstudio and the tidyverse library
Instructor(s):
Ekaterina Avershina (main)
Einar Birkeland, Trine B Rounge
Streamlining Bioinformatics Data Pipelines with Omnipy
Date: Thursday 12 December 2024 13:00-16:00 Add to calendar
Room: Prolog (room 2465)
Description
The time is ripe to challenge fundamental assumptions on managing bioinformatics data. With good reason, the notion of piping together command line tools as workflows has long dominated bioinformatics. However, returning to flat files on the command line at each junction adds parsing and serialisation steps that are unnecessary and often change the data in subtle (or less subtle) ways that are hard to fix. With Omnipy, we instead follow the lead of ETL (Extract, Transform and Load) solutions for Big Data – streamlined as an elegant and powerful Python library. Omnipy offers a systematic and scalable approach to building pipelines where the data is in focus, more than the tools. Through specification of data models/parsers, Omnipy allows researchers to import data in various formats and wrangle the data through stepwise transformations. For automation with large data, Omnipy seamlessly scales up for deployment on remote infrastructures. The workshop will introduce the Omnipy library and its main concepts. This will take form as a mix of presentations and hands-on exercises. While Omnipy is designed for cross-domain applicability, the primary use cases are bioinformatics-related. We will thus make use of real-world examples that should feel relevant to many of the attendees. We have held similar workshops before, e.g. for Oslo Bioinformatics Week 2023 (https://www.mn.uio.no/sbi/english/events/oslo-bioinformatics-workshop-week-2023/#02) and Digital Scholarship Days 2024 (https://www.ub.uio.no/english/courses-events/events/dsc/2024/digital-scholarship-days/22-omnipy-part1.html and https://www.ub.uio.no/english/courses-events/events/dsc/2024/digital-scholarship-days/22-omnipy-part2.html) at UiO. Continuously improved over several years – not the least through feedback from workshop attendees – Omnipy is finally getting ready for its v1.0 release!
Learning outcomes
- Introduction to Python type hints and Pydantic models
- How to use type hints to define models, datasets, tasks and flows in Omnipy
- How to write a simple parser for a tabular file format
- How to set up an executable mapping of data from one metadata schema to another
- How to automate an Omnipy data pipeline by deploying it to the Prefect orchestrator on NIRD (National Infrastructure for Research Data)
Target audience
PhDs, Postdocs, Technical personnel. Interest and experience with programming in general, and specifically data wrangling of bioinformatics data/metadata.
Pre-requisites
- The participants should have some experience with Python programming/scripting. We will not spend time explaining basic syntax and concepts, other than what is related to type hints. Experience with type hints in Python is useful, but not required.
Equipment to bring
Laptop, with an account set up for Google colab
Instructor(s):
Sveinung Gundersen (main)
Pável Vazquez
microRNA annotation and expression profiling
Date: Friday 13 December 2024 9:00-12:00 13:00-16:00 Add to calendar
Room: Python (room 2269)
Description
We will use state of the art microRNA annotation tools and reproduce a scientific paper
Learning outcomes
- Participants will be able to understand scientific papers on microRNA annotation and expression profiling. Participants will learn to retrieve smallRNAseq data from public repositories, process them and plot them.
Target audience
Master level and upwards
Pre-requisites
- bash and basic R
Equipment to bring
laptop with genuine commandline and administrator rights. We will send instructions for installation of software a week before
Instructor(s):
Bastian Fromm (main)
Alessandro Formaggioni
Deep Learning for Drug-Target Interaction Prediction and Drug Repurposing
Date: Tuesday 10 December 2024 9:00-11:00 Add to calendar
Room: Python (room 2269)
Description
This workshop will explore deep learning applications in drug discovery, focusing on two key areas: drug-target interaction (DTI) prediction and drug repurposing. Participants will learn how deep learning models, such as Convolutional Neural Networks (CNNs), can be utilized to predict interactions between drugs and targets, as well as identify potential repurposing opportunities for existing drugs. The workshop will include hands-on coding sessions using publicly available datasets, such as the KIBA dataset, and practical model implementation.
Learning outcomes
- After attending this workshop, participants will gain an understanding of how to use deep learning in two significant aspects of drug discovery: drug-target interaction prediction and drug repurposing. They will gain practical experience in data preparation, model building, and evaluation, providing them with the skills needed to leverage AI techniques in drug discovery and development.
Target audience
Early-stage researchers, graduate students, and professionals in computational biology and bioinformatics with a basic understanding of machine learning and Python.
Pre-requisites
The workshop's prerequisites include:
- Basic understanding of Python programming, e.g., familiarity with core Python data structures, such as lists, dictionaries, and NumPy arrays and experience with basic data manipulation libraries like pandas 2.A foundational understanding of supervised learning and how neural networks function at a basic level (e.g., concepts like layers, weights, and activation functions). Although advanced familiarity with deep learning frameworks is not required, some previous exposure to Pytorch is beneficial 3.Some experience with Google Colab (or at least Jupyter Notebook)
Equipment to bring
A laptop
Instructor(s):
Damla Övek Baydar (main)
Exploring Perturbation-Linked scRNA-Seq and GeneSetR for Functional Genomics Analysis
Date: Tuesday 10 December 2024 13:00-15:00 Add to calendar
Room: Python (room 2269)
Description
This workshop offers a comprehensive introduction to perturbation-linked single-cell RNA sequencing (scRNA-seq), a cutting-edge technique that combines genetic perturbations with single-cell transcriptomics to unravel gene functions and regulatory networks at an unprecedented resolution. Participants will learn about the principles of perturbation-linked scRNA-seq, its experimental workflow, and the advantages it offers over traditional methods, such as high-throughput functional screening and the ability to dissect complex biological systems at the single-cell level. Building on this foundation, the workshop will introduce GeneSetR, a user-friendly tool designed to harness the power of Genome-wide Perturb-Seq (GWPS) data for in-depth analysis of user-defined gene lists. GeneSetR empowers researchers to interpret and cluster gene lists effectively, facilitating navigation through the recently available GWPS datasets.
Learning outcomes
- Participants will gain an understanding of perturbation-linked scRNA-sequencing, including its methodology, applications, and advantages in functional genomics research. The workshop will cover how to use GeneSetR to find genes that regulate or are regulated by a specific gene of interest, analyze RNA-seq results to identify significant biological pathways, and cluster gene lists to identify functional groups. Attendees will learn to create correlation maps based on knockdown profiles and gene expression upon perturbations, as well as generate heatmaps illustrating relationships between chosen genes and perturbations. Additionally, the workshop will demonstrate how to find perturbations that modulate a given gene signature, which is crucial for understanding gene function and regulation.
Target audience
Researchers in the fields of genomics, molecular biology, and functional genomics, Bioinformaticians and Computational Biologists. Students who are interested in recent developments in functional genomics.
Pre-requisites
- Access to web.
Equipment to bring
none
Instructor(s):
Omer Faruk Kuzu (main)
Statistical principles in machine learning for small biomedical data
Date: Tuesday 10 December 2024 13:00-16:00 Add to calendar
Room: Prolog (room 2465)
Description
A central problem in machine learning is how to make an algorithm perform well not just on the training data, but also on new inputs. Many strategies in machine learning are explicitly designed to reduce this test error, possibly at the expense of increased training error. These strategies are collectively known as regularisation and they are instrumental for good performance of any kind of prediction or classification model, especially in the context of small data (many features, few samples). We will discuss basic connected concepts of generalisation, overfitting, bias-variance trade-off and regularisation and will illustrate the principles with penalised (generalised) linear regression models, with ridge, lasso and elastic net penalties as prominent examples. Finally, we will present the idea of structured penalties and priors, which can be tailored to account for structures present in the data, e.g. multi-modality or complex correlation structures. We will use examples from large-scale cancer pharmacogenomic screens, where penalised regression and alternative Bayesian approaches are used for predicting drug sensitivity and synergy based on the genomic characterisation of tumour samples. In the hands-on tutorial we will use R to perform an integrated analysis of multi-omics data with penalised regression.
Learning outcomes
- After attending this workshop, learners will understand key concepts for training machine learning models such as regularisation and how to incorporate data structure in the regularisation process.
Target audience
Graduate, post-graduate students and researchers at any level who are interested in applying machine learning methods to small data (i.e., few examples but potentially many features) or noisy data (e.g. biomedical data).
Pre-requisites
- Basic familiarity with R. Introductory level statistics including regression.
Equipment to bring
Laptop with a recent version of R and RStudio installed.
Instructor(s):
Manuela Zucknick (main)
Theophilus Asenso
There's a snake in my RNA-seq - Introductory course to single cells RNA-seq analysis with Python
Date: Thursday 12 December 2024 13:00-16:00 Add to calendar
Room: Python (room 2269)
Description
RNA-seq analysis is moving very fast towards the single-cell world. Luckily, the Python bioinformaticians community is moving accordingly fast, developing and supporting many tools to ease out the complicated tasks of sequencing analysis. This workshop will introduce the attendants to the basics of single-cell analysis using Python notebooks, scanpy, and the scverse. The workshop will introduce Scanpy, the core of the scverse, software designed to handle single-cell data with the flexibility of a Pandas data frame and some of the related packages developed and maintained by the community. We will go from raw data through QC, filtering, and dimensionality reduction. Spend some time learning why and how it is necessary to reduce the dimensionality of single-cell data and how to do that in a data-driven and information-driven manner. We will then proceed to the clustering phase and explain why understanding this process is fundamental to obtaining reliable and trustable results. Then we will conclude with the annotation phase, showing how we can use automatic methods as a starting point to obtain cell types and manually correct the labels based on our experience. If time permits I will show some packages to obtain Gene Regulatory Networks (GRN) and simulate perturbations on the cells by modifying the transcription of genes in the count matrix.
Learning outcomes
- The attendants will be able to use Scanpy as analysis software to process single-cell RNA-seq data. Perform QC, filtering, and clustering. Annotate the cell types with automatic methods and perform a data
- and experience-driven analysis.
Target audience
PhD students in Bio-related fields (molecular biology, genetics, etc) PhD students in Informatics who want to study some Bioinformatics analysis
Pre-requisites
- Knowledge of very basic Python Usage of Python Notebooks and installation via Pip and Conda https://scverse.org/packages/#ecosystem Packages required: Scanpy Pandas Seaborn Decoupler CellOracle
Equipment to bring
Laptop, access to computational server if possible (e.g., biotin)
Instructor(s):
Samuele Cancellieri (main)
Improving software quality in bioinformatics through testing
Date: Monday 09 December 2024 9:00-12:00 Add to calendar
Room: Python (room 2269)
Description
What is testing? What types of tests are we usually performing and what types are we not? We argue that bioinformaticians, especially tool developers, should integrate unit tests and automated tests into their development practices. In this workshop we show and practice how to write unit tests and make your code testable. Code testing is fundamental to achieve good quality software. Adding automated tests to a software base have several advantages. It increases confidence that the software performs as intended. Test coverage reduces the time required for ongoing development (eg. updating or adapting a tool or a pipeline) by ensuring that the new changes are not affecting the existing functionality in unexpected ways. Writing tests enforces thinking about cases where the code might fail, thus improving the usability of the software. This process also helps improve the modularization of the code, because complex, convoluted code is much harder to test. However, thinking about tests for the first time, especially thinking about good tests for testing scientific software is not easy. Through this workshop we aim to help you get started with unit testing. If time allows, we also spend some time on how to automate tests via GitHub actions.
Learning outcomes
- After attending this workshop, participants are more aware about the mindset and available tools to better test their software in an automated, detailed and thorough manner. We hope that participants will employ these techniques when they write their own tools and maybe even analysis pipelines, thus producing more robust and clean code.
Target audience
Bioinformaticians who write scripts, functions, or whole packages, but not quite (or not at all) get the hang of automated unit tests yet. Professionals of any career stages are welcome.
Pre-requisites
- Attendees should have some coding experience with Python or R.
Equipment to bring
Attendees should bring their laptops with set-up Python and/or R installations, so it is ready to be used. We encourage the participants to bring their own code (scripts or functions) that would be subjected to testing.
Instructor(s):
Katalin Ferenc (main)
Ieva Rauluseviciute, Ladislav Hovan