| layout | title |
|---|---|
page |
Past Events |
⬤ ELIXIR-Germany 2nd BioHackathon ⬤ 11-15 December 2023 ⬤ Bielefeld, Germany ⬤ Hybrid event ⬤
Spatially resolved transcriptomics allows us to resolve gene expression in the native tissue context. We are seeing an explosion in the number of tools being developed for downstream analysis and an emergence of datasets that are routinely used to evaluate the quality of new tools, however, we currently lack a community-driven continuous benchmarking of spatially resolved transcriptomics tools. A recent article in Nature highlights the need to address this issue via implementation of benchmarking via the OpenEBench and OMNIBENCHMARK platforms. These platforms provide a framework for curating computational tools and reference datasets for data, with a view towards extensibility as new computational approaches emerge.
The focus of our BioHackathon project will be on:
- computational tools to identify tissue domains and niches, commonly referred to as spatial clustering (e.g. SpaGCN, BayesSpace)
- reference datasets across technologies (e.g. Visum, Xenium, Slide-seq, MERFISH) and tissues (e.g. DLPFC, hippocampus, olfactory bulb, liver
- investigating evaluation metric (e.g. ARI, NMI) with a focus on novel metrics and qualitative interpretation
- community benchmarking via OMNIBENCHMARK
- generalisability of models/results across batches
While the primary aim of the project will be to have fun and meet other spatial researchers, we would endeavour to wrap up findings into publications. In the previous iteration of the SpaceHack project we attracted over 60 participants, and had a number of breakout projects that have preprints published on bioRxiv such as the SpatialData framework and the Xenium quality assessment study, with more on the way!
Target audience: PhD/postdoc-level bioinformaticians familiar with spatial omics analysis. We would strongly encourage those with prior experience in spatial clustering/domain identification to apply! This will not be a tutorial, so be ready to contribute ;p.
Dates: Monday 11th - Friday 15th December 2023
Location: Bielefelder-Hof Hotel, Bielefed, Germany. The venue for the BioHackathon Germany - Hotel Bielefelder Hof - is located right in the centre of Bielefeld, opposite the main train station and therefore offers an optimal connection and location. For transport recommendations please check the ELIXIR Germany event page.
Contact: for more information email Naveed Ishaque, [email protected]
Registration: registration is free using this webform before the deadline of 17.11.23. The number of on-site participants is limited to 80 persons.
Costs: registration for the BioHackathon is free, however, you will have to pay your own travel, food and accommodation expenses. de.NBI / ELIXIR Germany will pay expenses for lunch and drinks in the event.
Accomodation: please check the ELIXIR Germany event page for hotel recommendations.
Organizing committee:
- Naveed Ishaque, Berlin Institute of Health at the Charité, Germany. Naveed leads a bioinformatics research group with a strong focus on developing and applying computation methods to better understand spatially resolved transcriptomics data.
- Brian Long, Allen Institute for Brain Science, USA. Brian is a member of the Imaging Department at the Allen Institute with extensive experience in image and data analysis. He plays a driving role in the CZI funded SpaceTX consortia.
- Ahmed Mahfouz. Leiden UMC and TU Delft, Netherlands. Ahmed is an Assistant Professor leading a group that develops machine learning and statistical methods for high-throughput genomic data. They are interested in unraveling how genetic and environmental factors influence the molecular state of individual cells in normal and disease processes with applications in neuroscience and immunology.
- Mark Robinson. University of Zurich, Switzerland. Mark is a Professor leading a group that develops statistical methods for interpreting high-throughput sequencing and other genomics technologies in areas such as genome sequencing, gene expression, and regulation and analysis of epigenomes.
- Charlotte Soneson. Friedrich Miescher Institute for Biomedical Research, Switzerland. Charlotte is a research associate in the Computation Biology Platform at the FMI and is highly experienced in the analysis of transcriptomics data and benchmarking studies.
- Teresa Zulueta-Coarasa. EBI, UK. Teresa is a bioinformatician at the BioImage Archive and has multi-disciplinary experience in microscopy, quantitative image analysis, data management, and cell biology. Currently, she is working on improving open access to biological images and annotations to accelerate the development of AI methods.
- Christoph Kuppe. RWTH Aachen University, Germany. Christoph leads the lab of Quantitative Cell Dynamics and Systems Biology whose goal is to integrate computational tools and novel technologies to enable unprecedented insights into molecular signaling circuits of cells in healthy and diseased tissues.
- Michael Fletcher. Mike is a senior editor at Nature Genetics, Senior editor. Previously, Mike was a Postdoctoral Fellow in the Division of Molecular Genetics at the German National Cancer Research Centre (DKFZ). He ran computational analyses on high-throughput sequencing data generated for a variety of genomics projects studying the biology of adult glioblastoma, a late-stage brain cancer with an extremely poor prognosis for patients.
- George Gavriilidis. George is a pharmacologist/Post-Doc research working in the Biodata analysis group in INAB, CERTH (F. Psomopoulos Group) focusing on single-cell RNA analyses. He is also interested in Systems Immunology (maturation of B cells in health and disease), interpretable Machine Learning in scRNA-seq analysis and Systems Pharmacology approaches (hit identification, drug repurposing) based on single-cell multi-omics.
The ELIXIR Germany Code of Conduct has been created so all participants at ELIXIR Germany events organized can interact with each other in a respectful and safe environment, and have avenues to turn to in case they believe there has been a breach to this Code of Conduct. Be excellent to each other.
⬤ 12th-16th December 2022 ⬤ Leucorea, Lutherstadt Wittenberg, Germany ⬤ Hybrid event ⬤
Single cell omics has revolutionised the way and the level of resolution by which life science research is conducted, not only impacting our understanding of fundamental cell biology but also providing novel solutions in cutting-edge medical research [1,2]. More recently, several approaches have been developed to profile spatial single cell gene expression within the tissue context, providing maps of spatial locations and interactions of cell types [3]. Spatially resolved transcriptomics has been highlighted as an important technology for life science research through being awarded Nature Method of the Year 2020 [4]. Despite being in the limelight for 2 years, guidelines and workflows for best practices for the analysis of imaging-based spatially resolved transcriptomics data are somewhat lacking.
This project aims to combine the expertise of both biologically and technically focused researcheres to establish benchmarking datasets, comparing performance segmentation and cell assignment tools, and evaluating results in light of tissue specific challenges. A key foundation of the hackathon will be comparing spatial data to matched single cell data in the context of the ongoing TXsim project at the Helmholtz Centre Munich (Louis Kummerle, Malte Leucken and Fabian Theis). For the hackathon we will collate dataset of imaging-based spatial transcriptomics (e.g. Xenium/ISS, Vizgen/MERFISH, Resolve/MolecularCartography, seqFISH, EEL, etc) for a number of different organs (brain, lung, liver, heart) from both mouse and human to get things started.
Our primary focus areas for this workshop are:
-
Data. Identification and collation of datasets for benchmarking. Some key datasets will be collated and ready to use for the start of the hackathon, but we hope to expand this to include other relevant technologies and biological systems, especially with a focus on evaluating the quality of cell segmentation.
-
QC.Identification of key quality control metrics. We hope to establish clear definitions of key parameters for defining the quality and comparability of experiments such as number of UMI/molecules detected, off-target rate, consistency with prior knowledge from scRNAseq, noisy signals, sensitivity, etc.
-
Cell segmentation. Cells exhibit a diversity of morphologies, and the standard approach of segmenting cell nuclei/DAPI signal and expansion to model the cytoplasm doesnt always work, e.g. not all cells are uniformly round.
-
Normalisation and cell type annotation. Annotation of cell types in spatial data is still in it’s infancy, but there are some interesting paradigms emerging [e.g. 5].
-
Evaluation. results in light of experimental and tissue based consideration such as single-nuc vs single-cell RNAseq, cell density, tissue complexity, rare cells, non-uniformly shaped cells etc
References
[1] Tang et al (2009). “mRNA-Seq whole-transcriptome analysis of a single cell”. Nature Methods. https://doi.org/10.1038/nmeth.1315
[2] Aldridge & Teichmann (2021). “Single cell transcriptomics comes of age”. Nature Methods. https://doi.org/10.1038/s41467-020-18158-5
[3] Moses & Pachter (2022). “Museum of spatial transcriptomics”. Nature Methods. https://doi.org/10.1038/s41592-022-01409-2
[4] Marx (2021). “Method of the Year 2020: spatially resolved transcriptomics”. Nature Methods. https://doi.org/10.1038/s41592-020-01033-y
[5] Stringer et al (2021). “Cellpose: a generalist algorithm for cellular segmentation”. Nature Methods. https://doi.org/10.1038/s41592-020-01018-x
An important consideration for a successful hackathon would be community engagement. We plan to further enhance outreach to relevant communities including the ELIXIR single cell omics community, SCOG, the CZI SpaceTX consortia, and other key stakeholders in the field of spatial transcriptomics within Germany and internationally. We also plan to reach out to companies (e.g. Resolve Biosciences, 10x Genomics, Vizgen, Rebus, etc) to learn from their experiences and identify key areas that require attention. In line with Open Science principles, we would aim to share our findings with the wider scientific community – by providing good data and access to robust and easy to use tools in an accessible framework, would make life a lot easier for people new to the field. We also hope to wrap up our major findings in white papers and/or contributions to ongoing studies.
Organizers:
- Naveed Ishaque, Berlin Institute of Health at the Charité, Germany. Naveed leads a bioinformatics research group with a strong focus on developing and applying computation methods to better understand spatially resolved transcriptomics data.
- Brian Long, Allen Institute for Brain Science, USA. Brian is a member of the Imaging Department at the Allen Institute with extensive experience in image and data analysis. He plays a driving role in the CZI funded SpaceTX consortia.
- Louis Kuemmerle. Helmholtz Center Munich, Germany. PhD student in computational biology in the Theis and Erturk lab, with a strong focus on cell segmentation for spatially resolved transcriptomics data.
Target audience: PhD/postdoc-level bioinformaticians familiar with spatial omics analysis. We would strongly encourage those with expertise in tissue specific segmentation challenges and/or working experience of analysing spatial/single-cell data sets to apply!
Dates: Monday 12th - Friday 16th December 2022
Location: Lutherstadt Wittenberg - Leucorea
Contact: for more information email Naveed Ishaque, [email protected]
Registration: registration is free using this webform before the deadline of 25.11.22
Costs: registration for the BioHackathon is free, however you will have to pay your own travel, food and accommodation expenses. de.NBI / ELIXIR Germany will pay expenses for lunch and drinks in the Leucorea.
Accomodation: the event will be taking place in Wittenberg, Germany. for accomodation recommendations, please refer to the official event website.
The ELIXIR Germany Code of Conduct has been created so all participants at ELIXIR Germany events organized can interact with each other in a respectful and safe environment, and have avenues to turn to in case they believe there has been a breach to this Code of Conduct. Be excellent to each other.