Programme

Monday 10 Sep Tuesday 11 Sep Wednesday 12 Sep Thursday 13 Sep Friday 14 Sep
8:30–9:15 Registration
9:15–9:30 Welcome speech
9.30–11.00 Usadel Usadel Usadel Usadel Workshop

11.00–11.30

Coffee Break Coffee Break Coffee Break Coffee Break Workshop
11.30–13.00 Sanguinetti Sanguinetti Sanguinetti Sanguinetti Workshop
13.00–14.30 Lunch Lunch Lunch Lunch Workshop
14.30–16.00 Danos Danos Danos Danos Workshop
16.00–17.00 Coffee Break and poster session Coffee Break and poster session Coffee Break and poster session Coffee Break and poster session Workshop

COURSES

Main topic: biological interaction networks

Area: Mathematics
Lecturer: Dr. Guido Sanguinetti, University of Edinburgh, UK.
Abstract: Uncertainty is inherent in many aspects of biology, from the intrinsic noise of cellular reactions, to the extrinsic noise due to fluctuating environments, to inevitable experimental noise in the measurement process. Proper handling of uncertainty is essential in many steps of model development. In these lectures, I will review the mathematical foundations of stochastic modelling and introduce some more advanced tools for statistical inference in models of biological systems. I will introduce the basic concepts of probability theory and focus on Bayes' theorem as a tool for calibration and uncertainty quantification. I will explain some concepts of statistical inference such as Markov chain Monte Carlo and variational methods. I will then present some basic time-series models and their use in biology, and conclude discussing more advanced continuous time stochastic models.
Bio: Guido Sanguinetti received his degree in Physics from the University of Genova and his DPhil in Mathematics from the University of Oxford. He was a postdoc, then lecturer, in the Department of Computer Science at the University of Sheffield prior to joining the faculty at the School of Informatics, University of Edinburgh, in 2010. His interests focus on mathematical and statistical models of dynamic biological systems. 

Area: Computer Science
Lecturer: Prof. Vincent Danos, University of Edinburgh, UK.
Abstract: We will describe a new methodology to describe, simulate and investigate complex biomolecular networks. This method is called rule-based modelling and has the advantage that it can cope better with combinatorial molecular systems than usual reaction-based methods. The following aspects will be covered: knowledge representation, simulation, causal analysis, model reduction techniques.
Bio: Vincent Danos graduated in Engineering, obtained a PhD in maths. He has a 20 years academic career in logic and theoretical computer science, with an increasing concern for applications, mostly in formalising, modeling and analysing complex systems—e.g., biomolecular networks. 

Area: Biology
Lecturer: Prof. Bijoern Usadel, RWTH Aachen University, Gernamy
Abstract: The last decade has seen a massive explosion of omics data becoming available to the individual researcher. Initially, the focus was on individual experiments focusing on the limited study of a certain condition. However, given the massive growth of omics data in public data bases, these data can be holistically integrated and novel inferences made. In the beginning this was e.g. based on large scale approaches using simple correlation and a guilt by association approach, having lead to a massive knowledge gain for experimental biologists. Here we present several different streams of how to combine public (and own) dataset stemming from different disciplines in order to make new inferences about the plant as a whole. Firstly we present a novel normalization method for Affymetrix type microarrays beneficial for correlation analysis. We then show how this normalized transcript data from focused areas can be used to predict plant status which we validate using metabolite data. Based on these models we combined metabolite and transcript data sets to make informed decisions about gene knock-out experiments validating our predictions. We also show that guilt by association approaches can be improved by incorporating novel measures, if a large data set is properly mined using expert rules. These results also imply that the automatic incorporation of additional e.g. sequence information will aid in data interpretation. As a proof of concept we show that integration of sequence with simple microarray derived expression data leads to an improved predictor for plant protein chloroplast import. We finish by showing that next-generation sequencing data is not making matters more complicated but will allow us to get an even deeper understanding of living organisms.
Bio: Dr. Björn Usadel studied in Biochemistry in Berlin and New York where he worked in Prof Ulrike Gaul's lab on the development on the visual system of Drosophila. During this time he got interested in Bioinformatics and then went on to Golm where he did his PhD in the group of Dr. Markus Pauly on the identification and characterization of novel cell wall genes. He then worked as a Postdoc in Prof. Mark Stitt's lab on the visualization and evaluation of high throughput data. Since 2008 he was a group leader at the Max Planck Institute and since then worked on data visualization, analysis as well as sugar status and cell wall biosynthesis. Since 2011 he is a full pofessor at the RWTH Aachen university and a co-director at the Forschungszentrum Jülich.