1st Unseminar: Challenges in Systems Biology

From HUB
Jump to navigation Jump to search

What? Meet fellow bioinformaticians in Heidelberg, and share tips and tricks for overcoming Challenges in Systems Biology

When? Thursday April 26, 2012, 19.00 - 21.00

Where? Grabengasse 3-5, Building 2170, Groundfloor, Hörsaal 02.

How? The 1st HUB will feature talks and discussions, but you get to shape the programme.

Who is organising it? A team of people working at several different life-science organisations within Heidelberg are organising the 1st unseminar, with your help.

Who is coming? Check out the list of registered participants

Registrations and proposals for talks and discussion sessions are now closed

Results of survey from 1st unseminar


19:00 - Welcome

19:05 - Introductions

19:20 - Introductory talk: From Bench to Website - Thomas Lemberger, Chief Editor, Molecular Systems Biology

19:35 - Discussion: Biologists and Bioinformaticians: Building Successful Collaborations (Moderated by Raeka Aiyar)

20:00 - Discussion: Systems Biology and Human Diseases, including two flash talks selected from those proposed by you (Moderated by Kallia Trachana)

20:35 - Ideas for the next meeting

When we're finished - Off to the pub: Bier Brezel (Map)

Your contributions to shaping the programme

Thanks to everyone for all their ideas and enthusiasm. Unfortunately there won't be enough time to cover everything but we hope to come back to these another time.

Proposed Discussion Topics

Always wondered how others deal with a challenge you frequently face? Have a proven approach to a problem, or know someone who does? This is the place to put forward what keeps you up at night, and what gets you out of bed in the mornings.
Suggest a topic and a good person to moderate the discussion (this can be you, of course), or add your thoughts on the current suggestions:

Biologists and Bioinformaticians: Building Successful Collaborations

Your strategies and experiences for a successful collaborations between bench biologists and bioinformaticians in systems biology

Goals: Please contribute a two sentence statement of what you would like participants to get out of the discussion.

Main Discussion Points

  • Specific experiences participants have had with such collaborations
    • features that made them successful
    • features that made them less successful
    • signs they noticed that suggested it was going to go well or badly
    • successful strategies they used to make such projects work well i.e. what actions can we take to make these a success

Proposed Structure

  • 2 minute intro to the topic by the moderator
  • Invite the audience to spend 3 or 4 minutes discussing the main topic points with their neighbors
  • Open discussion of main topic points
    • White boards or paper charts used to collect these i.e. participants go and write on these their ideas
    • Ask members of the audience to share the ideas with the audience about these different issues we've discussed
    • Moderator (or someone working with them) keeps a list of these
  • Near the end of the session, the list of ideas is shown on a screen, and we vote to decide which ones we find most important

Systems Biology and Human Diseases

In general, this section can be dedicated to personalize medicine and how this is possible through sequencing and other omic data analyses. At least two flash talks (Mani and Julien) are related to this topic (from a different aspect however). I think that Korbel lab should be involved in this topic (they have multiple successful stories on it).

  • A systems biology approach for personalized/stratified medicine
    • disease(patient)-stratification
      • cancer (Korbel lab)
      • chronic diseases (e.g. diabetes) (Mani)
    • response-stratification using an alternative model organism (Julien)

Recurring Challenges in Systems Biology Data

Large-datasets, format standards, data exchange, assessing significance and validation

Goals: Please contribute a two sentence statement of what you would like participants to get out of the discussion.

  • Proposed by Aidan Budd
  • No suggestion yet for a Moderator

Main Discussion Points

  • Particular challenge participants have found when working with systems biology data (if they have no experience of such work, then they should think about what they would expect the major challenges are):
    • List potential discussion points here

Kalliopi would separate the listed challenges into:

  • large scale, tool development and storage; I guess tool developers for large datasets are qualified to talk (Jens/Shini from our group or Alexi's lab for RaxML) or administrators (e.g. Yan from our lab). I would also propose Maria Secrier, as a speaker, as she is working on the visualization of large data (quite important problem) and a quite nice aspect of biological systems (temporal regulation).
  • large scale and complexity; how we deal with biological complexity (integrating omic data, increasing the studied samples, using evolutionary information etc) and how we can translate the complexity into functionality (which is a good transition about translational informatics and personalize medicine)

Better use of kinetic parameters in Systems Biology


  • Proposed by Jon Fuller
  • I could moderate the session?

Main Discussion Points

  • What experimental data is currently missing, how to fill this gap?
  • More details to follow...

Proposed Flash Talks


Human gut microbiome and stratified medicine

  • Mani Arumugam
  • EMBL-Heidelberg
  • The human gut microbiome is an integral and largely ignored part of the human body. It constantly interacts with the immune system and is associated with several diseases. Recently we discovered that people can be stratified into groups based on their microbiome composition. This may have an impact on stratified medicine, bringing in a human microbiome component to stratified medicine.

A mathematical model of hindgut curvature

  • Joseph Barry
  • EMBL-Heidelberg
  • A novel mechanism for the maintenance of hindgut curvature in the developing Drosophila embryo is explained in the context of intercellular adhesion and the differential interfacial tension hypothesis.

Sequencing of transcript boundaries reveals unseen transcriptome complexity

  • Aino Järvelin
  • EMBL-Heidelberg
  • Variation in transcript isoforms influences the functional output of the genome. We have recently developed RNA-sequencing based technologies to better characterize the architecture of transcriptomes. Our approach reveals extensive (same strand) overlap of transcripts in the yeast genome, frequent usage of multiple start and end sites, and bidirectionality of polyadenylation sites of convergent transcripts. Our current and future goals include gaining a better understanding of the regulation and consequences of transcriptome complexity.

Negative protein-protein interaction datasets derived from large-scale two-hybrid experiments

  • Leonardo Trabuco
  • University of Heidelberg
  • Negative protein-protein interaction datasets are needed for training and evaluation of interaction prediction methods, as well as validation of high-throughput interaction discovery experiments. We present a simple method to harness two-hybrid data to obtain negative protein-protein interaction datasets, validated using available experimental data.

Adding details to interaction networks

  • Matthew Betts
  • CellNetworks / University of Heidelberg
  • Many experimental methods for detecting protein-protein interactions tell us what interacts with what but not how. Protein 3D structure can provide these details in many cases. I'll explain how we use protein structure to explain interactions and to predict new ones, and how we use it to investigate the impact of mutations and post-translational modifications.

Joint models of gene expression and global phenotypes in the context of genetic and environmental variations

  • Julien Gagneur
  • EMBL-Heidelberg
  • Joint models of gene expression and global phenotypes in the context of genetic and environmental variations
  • Phenotypic traits such as disease susceptibility are controlled by the joint influence of genotype and environment. Dissecting the molecular mechanisms that underlie these dependencies promises to deliver the necessary insights to develop drugs tailored to the genetic background and life circumstances of the patient. To this end, we have combined cellular and molecular phenotyping with statistical modelling to explain the molecular state as a function of genotype and environment and to provide for specific predictions of molecular intervention points.

Cell Tracking in Multicellular Environment

  • Kota Miura
  • EMBL-Heidelberg
  • Measurement of when, where and how cells are migrating/moving/changing shape is fundamental parameters of multicellular development and homeostasis, and are challenging tasks in image analysis of biological samples. I will present a brief overview on how we are working it (of course, the challenge is still going on).

Understanding long-term evolution of cell types: insights from comparative genomics, population transcriptomics, and developmental biology

  • Oleg Simakov
  • EMBL-Heidelberg
  • We are interested in reconstructing the functional cell type repertoire of the bilaterian ancestor and study its subsequent diversification. Previous work based mostly on parsimony is however undermined by the lack of knowledge of small-scale (environmental and population) and long-scale (genomic) variation that have played a major role over the last 600 million years. Here I discuss a new integrative approach that combines all this information and helps to shed light on multi-scale correlates of slow and fast-evolving cell types, thus allowing to design more solid evolutionary models.

The Virtual Pancreatic Beta Cell: A Modeling and Simulation Approach in Diabetes Drug Discovery

  • Chris Schneider
  • Sanofi-Aventis Pharma, Frankfurt
  • In the context of diabetes drug discovery we have developed a mathematical model of insulin secretion in the pancreatic beta cell. This was done by using published mathematical models that cover sub-parts of this biological system. Using this modular approach we are able to simulate glucose-stimulated insulin secretion and the amplification via GPCR activation.

Fitting high throughput data to a Boolean network model

  • Guy Karlebach
  • DKFZ - Heidelberg
  • Fitting between a high throughout dataset and a Boolean network model is an interesting problem with practical implications. In this flash talk I define the problem, analyze its complexity and present an algorithm for it.