The QBiC tech-talks focus on technical aspects in quantitative biology. They are providing insights in innovative technologies, methods and applications in life science. Experts with an intensive research background and/or are have their focus on services in quantitative biology will present the modern fields in quantitative biology.
- Dr. Gary Bader, 05.11.2020, 15:00-16:00, Online via Zoom and
Youtube (+ option to talk with the speaker afterwards): https://youtu.be/LMT7xY6KsAg, "Systems biology from single cell genomics to network data integration"
Single-cell RNA sequencing (scRNAseq) can map cell types, states and transitions during dynamic biological processes such as development and regeneration. Many trajectory inference methods have been developed to order cells by their progression through a dynamic process. However, when time series data is available, these methods do not consider the available time information when ordering cells and are instead designed to work only on a single scRNAseq data snapshot. We present Tempora, a novel cell trajectory inference method that orders cells using time information from time-series scRNAseq data. In performance comparison tests, Tempora accurately inferred developmental lineages in human skeletal myoblast differentiation and murine cerebral cortex development, beating state of the art methods. Tempora uses biological pathway information to help identify cell type relationships and can identifyimportant time-dependent pathways to help interpret the inferred trajectory. Our results demonstrate the utility of time information to supervise trajectory inference for scRNA-seq based analysis.
Biological networks have the power to map cellular function, but only when unified to overcome their individual limitations such as bias and noise. Unsupervised network integration addresses this, automatically weighting input information to obtain an accurate, unified result. However, existing unsupervised network integration methods do not adequately scale to the number of nodes and networks present in genome-scale data and do not handle frequently encountered data characteristics (e.g. partial network overlap). To address this, we have developed an unsupervised deep learning-based network integration algorithm that incorporates recent advances in reasoning over unstructured data – namely the Graph Convolutional Network (GCN) – that can effectively learn dependencies between physical, co-expression and genetic interaction network topologies. Our method, BIONIC (Biological Network Integration using Convolutions), produces high quality gene and protein features which capture and unify information across many diverse functional interaction networks. BIONIC learns features which contain substantially more functional information compared to existing approaches, linking genes and proteins that share co-complex, pathway and bioprocess relationships.
- Dr. Kenneth Berendzen, 08.10.2020, 16:00-17:00, Online via Zoom and
Youtube (+ option to talk with the speaker afterwards): https://youtu.be/04UiqvgZ9R0, "Using flow cytometry in plant science: advantages and limitations"
Flow cytometry is a technique whereby cellular components and states can be measured directly in whole cells using fluorescence techniques. By nature, flow cytometry consists of multi-dimensional data which is traditionally broken down into two-dimensional bivariate plots. While flow cytometry has been applied to plant science since its inception, its use is still not as widespread as that in animal science. This talk aims at introducing flow cytometry and its applications routinely used in plant sciences.
- Dr. Evangelia Petsalaki (EMBL-EBI, UK), 10.09.2020, 16:00-17:00, Online via Zoom and YouTube (+ option to talk with the speaker afterwards): https://youtu.be/Kmx69gBAD9g, "Integrative studies of context-specific signalling"
Our group aims to understand and describe the organisation principles of cell signalling that allow the diverse and context-specific cell responses and phenotypes. It is well established that signalling responses happen through complex networks. However, most signalling research still uses linear pathways as the ground truth. Moreover, signalling responses are highly dependent on context, such as tissue type, genetic background etc and therefore these static pathways are not always suitable. There is also a high bias in the literature towards kinases and pathways for which reagents and prior knowledge is readily available. This leaves a huge dark space in our understanding of cell signalling and significantly hinders studies of its general principles. Data-driven methods to study cell signalling using phosphoproteomics have typically been developed on transcriptomics data. Due to fundamental differences of these data types, including the sparsity of phosphoproteomics, these methods perform poorly on signalling network inference. Finally, large-scale essentiality datasets that can shed light on the context in which different signalling-related genes and pathways are crucial have thus far been severely underused in largescale omics data integrative efforts, likely due to their inaccessibility to non-expert users.
In this talk I will present two projects where we try to mitigate some of the above issues. For the first one I will present a method for data-driven machine-learning-based approach that takes advantage of global phosphoproteomics datasets to predict kinase regulatory networks, including their direction and sign of regulation. Our predicted kinase regulatory network is able to recapitulate known signalling pathways and agrees with orthogonal validation datasets. We were able to provide predictions for a large fraction of the understudied kinase space and found that kinase regulatory networks are denser than previously suspected.
For the second one I will present CEN-tools, an integrative webserver and python package, that allows users to navigate the contexts of different gene essentialities. I will demonstrate examples of its use in discovering new gene-gene relationships and important putative signalling targets for different cancers.
- Natalie Krieger, Sven zur Oven-Krockhaus, 04.06.2020, 16:00-17:00, Online via Zoom and YouTube (+ option to talk with the speakers afterwards), "Application of optical microscopy methods in plant biology"
The visibility of organs, cells, subcellular structures, or even the subcellular localization of proteins is becoming more and more important in cell biology and can be achieved by the use of different microscopes and microscopy techniques. In contrast to mammalian cells, living plant cells possess subcellular structures causing strong autofluorescence background and make the spectral separation of the fluorophore of interest challenging.
To analyze the variety of plant samples, there are, in addition to “common” widefield microscopes, three confocal laser scanning microscopes available at the Centre for Molecular Biology of Plants (ZMBP) at the University of Tübingen. These confocal laser scanning microscopes (CLSM) allow higher resolution not only of fixed but also of living cells. Equipped with add-ons, these CLSM allow a subset of spectro-microscopy techniques based on fluorescence to analyze the interaction behavior of proteins/molecules in living plants. Furthermore, the usage of deconvolution software and/or specialized detectors allow the visualization of samples beyond the resolution limit.
In addition to these commercial instruments, a self-built microscope for special techniques is also available, which was designed and built in cooperation with the Institute for Theoretical and Physical Chemistry. On the one hand, this has the advantage that new or modified techniques can be quickly tested and evaluated, and on the other hand, the microscopy methods can be adapted to the specific requirements of plant samples.
Our research with this instrument currently focuses on the use of so-called super-resolution techniques to visualize structures beyond the optical diffraction limit (200 nm). In particular, we are investigating the dynamics of receptor complexes in living plant membranes, which will be demonstrated with a few examples.
The evaluation of the large data sets generated by super-resolution techniques and the optimization of measurement parameters can significantly benefit from the use of modern computational strategies, including deep-learning-based frameworks and machine learning. This synergy will likely play an increasingly important role not only here, but in all biomedical imaging procedures in the future.
- Jerven Bolleman, 20.05.2020, 15:00-16:00, Online via Zoom and YouTube (+ option to talk with the speaker afterwards): https://youtu.be/zfBhRqEr1h8, "Interoperabilty in the sciences: it is in the semantics"
FAIR you should have heard of it by now ;) Findable, Accessible, Interoperable and Reusable. Lofty goals, which sound grandiose, but are actually grounded in technical realities. The SIB Swiss Institute of Bioinformatics, as a confederation of independent research and service groups covering wide areas of interests, is always trying to meet those goals. The pragmatic bottom up choice by a number of SIB groups is to use RDF and SPARQL to achieve more Interoperable and Reusable data resources. RDF is a knowledge representation framework, while SPARQL is a graph query language that allows for federated querying. There will be a demonstration using Rhea, Hamap and UniProt as examples of how this plays out in practice. With a small hint on how these resources are interoperable with data sources outside of the life sciences such as the European Patent Office datasets.Many of the technologies we choose are great at answering certain questions, yet are extremely inflexible. Leading to high (re)development costs and too little data integration. This is not required and we can do better together. An example is SpOdgi, which takes a highly compressed Odgi genome variation graph and presents it as a SPARQL/rdf graph database in less than 900 lines of code. SpOdgi is used to link 1000+ SARS-CoV-2 genomes with UniProt, Bgee, Oma, Rhea, and NeXtProt into one knowledge graph at https://covid-19-sparql.expasy.org/.
- Dr. Benjamin Schubert, 06.02.2020, 16:00-17:00, Lecture hall: 7E02, Auf der Morgenstelle 16, 72076 Tübingen
"Computer-aided Vaccine Design: How Algorithms can help during the Selection and Assembly of Epitope-based Vaccines"
Dr. Markus List, 12.12.2019, 16:00-17:00, Lecture hall: N2, Auf der Morgenstelle 16, 72076 Tübingen
"Analysing large-scale epigenomic data"
Large amounts of epigenomic data are publicly available, yet their retrieval for downstream analysis is a research bottleneck. Typically, users download huge files that span the entire genome, even if they are only interested in a small subset (e.g. promoter regions) or an aggregation thereof. Moreover, complex operations on genome-level data are not always feasible on a local computer due to resource limitations. The DeepBlue Epigenomic Data Server mitigates this issue by providing a powerful programmatic API that can be accessed by data analysts and software tool developers. To optimally support biomedical researchers, we also provide DeepBlueR, an extensively documented Bioconductor R package and present typical analysis work-flows. While a majority of DNA methylation based studies focus on average methylation levels, we also show that bisulfite sequencing data uniquely enables us to capture intra-sample DNA methylation heterogeneity.
Dr. Angel Angelov, 14.11.2019, 16:00-17:00, Lecture hall: N2, Auf der Morgenstelle 16, 72076 Tübingen
"Basecaller accuracy in Nanopore sequencing"
In Nanopore sequencing, basecalling is the computational process of translating raw electrical signal (squiggle) to nucleotide sequence. Basecalling is therefore a key factor determining usability and acceptance of Nanopore sequencing in the scientific community. Oxford Nanopore Technologies (ONT) have recently released a new basecalling algorithm (called flip-flop), which promised a much better read accuracy than the default one.
In this talk, I will present a comparison of the basecalling accuracy of the default and the flip-flop algorithms, using a real-life bacterial genome sequencing dataset from our lab. The implications of using better basecallers on the accuracy of bacterial genome assemblies will be discussed, as well as the perspectives of using Nanopore-only data to obtain near-complete, error-free bacterial whole genome sequences.
Prof. Dr. Chris-Carolin Schön, 24.01.2019, 16:15-17:15, Lecutre hall: 7E02, Auf der Morgenstelle 16, 72076 Tübingen
"Accuracy of genomic prediction in structured plant populations"
Most traits of importance in plant and livestock breeding are regulated by many genes and follow a quantitative distribution. The assessment of these quantitative traits in performance tests is costly and time consuming. Thus, prediction of the genetic potential of individuals from their DNA sequence is highly desirable.
This talk will present advancements in genome-based prediction using genomic, phenotypic and genealogical data. Statistical models for prediction of genetic values and phenotypes will be introduced. I will discuss challenges arising from the high- dimensional nature of genomic information as well as from the large genetic diversity and genome complexity of maize. The effect of integrating knowledge on marker-trait associations from functional or QTL studies and the role of feature selection for improving prediction accuracy will be shown. Factors driving prediction accuracy besides ancestral relatedness will be presented for traits of different genetic architecture. How to make best use of the massive collections of genomic and phenotypic data not only for prediction of phenotypes but also to understand the genetic architecture of complex traits and genetic phenomena such as epistasis and pleiotropy will be discussed.
Dr. Phil Ewels, 08.03.2018, 16:00-17:00, Lecture hall: N2, Auf der Morgenstelle 16, 72076 Tübingen, “Standardising Swedish genomics analyses at SciLifeLab NGI”
The SciLifeLab National Genomics Infrastructure is one of the largest sequencing facilities in Europe. We are an accredited facility providing library preparation, sequencing, basic analysis and quality control for Swedish research groups. Our sample throughput requires a highly automated and robust bioinformatics platform.
In this talk I will describe how we handle data flow, analysis and quality control. I'll describe our work with tools such as Nextflow and MultiQC; some of our pipeline projects for Cancer analysis and RNA; and touch on some upcoming projects such as MegaQC and nf-core.
Holger Gantikow, 20.09.2018, 16:00-17:00, Seminar room D4A19, D-Bau, Auf der Morgenstelle 16, 72076 Tübingen, “Linux Containers for HPC - Container Technology/Engine Architecture 101”
Container technology has not only spread in "normal" IT in recent years, but has also established itself as an important basic component in the field of High Performance Computing (HPC). However, here Docker is not the omnipresent solution, but there are a number of alternatives that can often be better integrated into existing HPC environments. The presentation will introduce which problems of HPC can be mitigated with containers and will refresh which technologies are based on containers.
Dr. Thomas Soddemann, 25.02.2016, 16:00-17:00, Lecture hall: N1, Auf der Morgenstelle 16, 72076 Tübingen, “Compute Intensive Applications: fast everywhere” (Fraunhofer Institute for Algorithms and Scientific Computing SCAI)
Dr. Dieter Beule, 03.03.2016, 16:00-17:00, Lecture hall: N1, Auf der Morgenstelle 16, 72076 Tübingen, "Tools Selection for Neo-Epitope Selection" (Berlin Institute of Health)
Prof. Dr. Frank Oliver Glöckner, 17.08.2016, 11:00-13:00, Lecture hall: N1, Auf der Morgenstelle 16, 72076 Tübingen, "The German Federation for Biological Data (GFBio): Services for Data Management and Long-term Archiving” (Head of the Microbial Genomics and Bioinformatics Research Group, Max Planck Institute for Marine Microbiology, Jacobs University Bremen gGmbH)
Prof. Guowang Xu, 09.04.2015, 16:00-17:00, Lecture hall: N1, Auf der Morgenstelle 16, 72076 Tübingen, "Liquid chromatography – mass spectrometry methods for lipidomics” (Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, China)
Dr. Nico Pfeiffer, 30.06.2015, 16:00-17:00, Lecture hall: N1, Auf der Morgenstelle 16, 72076 Tübingen, “Machine Learning for Personalized Medicine: How to Integrate and Interpret Data from Different Molecular Measurements” (MPI für Informatik, Saarbrücken)
Dr. Anne Zeck, 12.11.2015, 16:00-17:00, Lecture hall: N1, Auf der Morgenstelle 16, 72076 Tübingen, "Mass spectrometry of intact proteins for quality control and biological function elucidation” (Natural and Medical Sciences Institute (NMI), Reutlingen)
Dr. Thomas Joos, 01.01.2014, 16:00-17:00, Lecture hall: N2, Auf der Morgenstelle 16, 72076 Tübingen, "Immunoassays in Multiplex for Biomarker Discovery and Validation” (Deputy Managing Director, NMI Natural and Medical Sciences Institute at the University of Tübingen)
Prof. Rainer Bischoff, 22.05.2014, 16:00-17:00, Lecture hall: N1, Auf der Morgenstelle 16, 72076 Tübingen, “Biomarker Discovery and Validation” (University of Groningen, Department of Pharmacy, Analytical Biochemistry, 9713 AV Groningen, The Netherlands)
Prof. Dr. Christoph Borchers, 15.08.2014, 16:00-17:00, NMI Reutlingen, Markwiesenstraße 55, 72770 Reutlingen, "MRM and iMALDI, two emerging mass spectrometric approaches for biomedical and clinical research“ (University of Victoria, BC, Canada)
Dr. Oliver Schilling, 11.12.2014, 16:00-17:00, Lecture hall: N1, Auf der Morgenstelle 16, 72076 Tübingen, "How to use proteomics to investigate protease substrates and specificity"
& Dr. Lars Nilse, “Reliable Peptide Quantification” (Institute of Molecular Medicine and Cell Research, University of Freiburg)
- Prof. Detlef Weigel, 18.04.2013, 16:00-17:00, Lecture hall: N1, Auf der Morgenstelle 16, 72076 Tübingen, "Next generation genetics enabled by high-throughput sequencing technologies" (Max-Planck-Institute for Developmental Biology, Tübingen)
Prof. Boris Macek, 25.04.2013, 16:00-17:00, Lecture hall: N1, Auf der Morgenstelle 16, 72076 Tübingen, "Global detection of kinase substrates using quantitative mass spectrometry-based proteomics" (Proteome Center Tübingen, University of Tübingen)
Prof. Peter Bauer, 02.05.2013, 16:00-17:00, Lecture hall: N2, Auf der Morgenstelle 16, 72076 Tübingen, “Next generation sequencing goes diagnostics" (Department of Medical Genetics, University of Tübingen)
Prof. Olaf Riess
Prof. Dr. Marius Ueffing, 06.06.2013, 16:00-17:00, Lecture hall: N2, Auf der Morgenstelle 16, 72076 Tübingen, "Molecular dissection and scaled integration of functional protein networks: analyzing the molecular basis of human vision” (Institute for Ophtalmic Research, University of Tübingen)
Dr. Marc Stahl, 04.07.2013, 16:00-17:00, Lecture hall: N1, Auf der Morgenstelle 16, 72076 Tübingen, "Metabolomics: What we can do – what we cannot do” (Center for Plant Molecular Biology, University of Tübingen)
Prof. Peter Martus, 25.07.2013, 16:00-17:00, Lecture hall: N1, Auf der Morgenstelle 16, 72076 Tübingen, “Multiple Imputation of Missing Values – strategies and pitfalls” (Institute for Clinical Epidemiology and Applied Biometry, University Hospital Tübingen)
Dr. Oliver Poetz, 24.10.2013, 16:00-17:00, Lecture hall: N1, Auf der Morgenstelle 16, 72076 Tübingen, "Trap & Measure: Immunoassays In Targeted and Non-Targeted Proteomics” (NMI Natural and Medical Sciences Institute at the University of Tübingen)
Dr. Dr. Saskia Biskup, 07.11.2013, 16:00-17:00, Lecture hall: N1, Auf der Morgenstelle 16, 72076 Tübingen, "Next generation sequencing in human genetic diagnostics” (CeGaT Tübingen)
The Workshop will comprise lectures and relevant applications in the field of "Bioinformatics in Cancer Research”. Several distinguished guest speakers as well as experts from QBiC will share their research at various fronts in the field of cancer research.
26.11.2020, 10 a.m. to 4 p.m. via Zoom
PD Dr. med. Juliane Walz (KKE Translational Immunology)
Dr. med. Markus Löffler (Department of General, Visceral and Transplant Surgery)
Dr. Nicolas Casadei (Institute of Medical Genetics and Applied Genomics)
Dr. Andre F. Martins (Werner Siemens Imaging Center)
The aim of this workshop is to present a wide range of novel bioinformatics methodologies in cancer research and to create a platform to interact with experts in their respective fields.
Major breakthroughs in this field during the past decade have revolutionized cancer research, allowing omics and biomedical imaging data to be generated at ever increasing volumes and resolution. The challenge however is the increasing complexity in handling the resulting datasets which can be large and highly heterogeneous.
We will introduce computational methods for management, processing and analysis of complex omics and imaging data that are currently available at the Tübingen Life Science Campus and discuss how they are applied to cancer research.
|10:00-10:10||Dr. Sven Nahnsen|| |
Introduction to the workshop
|10:10-10:45||PD Dr. med. Juliane Walz||Peptide-based Immunotherapy - Antigens, Combinations, Translation|
|10:50-11:25||Dr. med. Markus Löffler||Cancer Immunotherapy: From Standard Therapy to Individualized Medicine|
|13:15-13:50||Friederike Hanssen||Bioinformatic analyses with portable and reproducible pipelines on public datasets|
|13:50-14:25||Dr. Nicolas Casadei|| |
Overview of methods used in molecular oncology
(DNA & RNA sequencing, ATAC-seq, bulk vs single-cell, methylation, short vs long reads).
|14:40-15:15||Dr. Andre F. Martins||Imaging tissue physiology with non-invasive metabolic approaches|
|15:15-15:45||Dr. Luis Kuhn||Semantic segmentation for medical image analysis|
|15:45-15:50||Dr. Sven Nahnsen||Conclusion|
|15:50-16:30||Post workshop breakout sessions: |
The development of reproducible, portable workflows is playing an increasingly important role in day-to-day data analysis at QBiC. The technological requirements for such workflows and analysis methods are reflected in the development of various domain-specific languages, which increasingly play a role in the scientific analysis of large data. At QBiC, the Nextflow platform was chosen for this purpose, which in combination with other projects such as nf-core has led to significantly improved workflows and standardization in the area of pipeline development.
In order to reinforce this development, a workshop with the core developers of Nextflow (Paolo Di Tommaso and Evan Floden) was initiated, which should offer beginners as well as advanced users the possibility to launch their own projects and pipeline developments with Nextflow.
The workshop will take place on April 11th/12th, 2019 in the premises of QBiC at the Tübingen University Hospital.
Registration is closed since March 29th. You will receive a bill with your institutional address that needs to be paid April 15th, 2019.
The workshop will be Thursday and Friday and the registration fee includes food and speaker fees.
We will update the agenda from time to time to make sure that everything is up to date. Also, there will be a social programme on Thursday, April 11 in the evening, so make sure to stay after the workshop in the evening!
Thursday, April 11
|09:00-09:30||Intro to Nextflow programming model|
|09:30-10:00||Get Started: First Nextflow pipeline|
|10:00-10:30||Basic language structures and commands|
|10:45-11:00||Basic language structures and commands continued|
|11:00-12:00||Nextflow Channels, concepts and usage|
|12:00-13:00||Nextflow Operators, concepts and usage|
Lunch (sponsored by AWS)
|14:00-15:45||Nextflow Processes, concepts and usage|
|16:00-17:00||A simple RNAseq pipeline|
|open||Try & Learn|
Social Dinner at 07:00 PM at Restaurant Neckarmüller.
Friday, April 12
|09:30-10:30||Managing dependencies (conda & containers)|
|10:45-11:30||Deployment scenarios (clusters & cloud)|
Error handling & troubleshooting
|12:15-13:00||Common implementation patterns|
|13:00-14:00||Lunch (sponsored by AWS)|
|14:00-17:00||Hands-on: Implementation Variant Calling Pipeline|
|16:00-17:00||Hands-on: Implementation Variant Calling Pipeline (continued)|