Huber Group
Computational Biology from Genomes to Phenotypes
Genomewide phenotypic map. a. Each of the 1,839 nodes represents an siRNA perturbation of cultured cells whose shape and morphology was monitored by automated microscopy. The graph is a two-dimensional representation of phenotypic diversity and similarity. b. Representative images for four siRNA perturbations. Cells were stained at the nuclei (DAPI, blue), actin (red) and tubulin (green).
Visualisation of genomic data with the Hilbert curve (S. Anders, Bioinformatics 2009)
Time courses of cellular states observed by live cell imaging are mathematically modeled by a dynamical system with differential equations of motion (Neumann et al., Nature 2010)
Independent filtering increases detection power for high-throughput experiments (Bourgon et al., PNAS 2010)
Previous and current research
The group studies genotypes and phenotypes on a genome-wide scale: how do variations in the genomes of individuals shape their phenotypes? To this end, we develop computational methods in statistics, probabilistic modeling, image analysis and bioinformatics.We work with experimental labs in systems genetics and functional genomics to design and analyse genome-wide experiments, with the aim of unravelling the mechanisms of genetic inheritance, gene expression, molecular interactions, signal transduction and how they shape phenotypes. Most phenotypes, including many human diseases, are governed by large sets of genes and regulatory elements interacting in complex, combinatorial networks. Our aim is to map and quantitatively understand these complex systems; and, eventually, to devise strategies for engineering phenotypes.
Our research is stimulated by new technologies, and we employ data from high-throughput sequencing (RNA-seq, ChIP-seq, genotyping, polymorphism discovery), large scale cell based assays, automated microscopy, as well as the most advanced methods of computational statistics.We are a regular contributor to the Bioconductor project (www.bioconductor.org).
Future projects and goals
One of the most exciting questions in biology is the predictive modelling of phenotypic outcomes based on individual genomes. To get there, we need a better understanding of the spectrum of genetic variations in a species, and how, combinatorially and together with environmental variations, they affect phenotype.
Progress in biology will continue to be driven by advances in technology. Sequencing will allow us to know the genomes of each individual person and model organism, and will provide transcription and DNA-protein interaction data for many different cellular systems at unprecedented depth. Light microscopy of single, live cells will provide data on molecular interactions and life-cycles within the cell. We will be able to observe the dynamics of signaling, the cell cycle, and cell migration both under normal conditions and many different perturbations (RNAi, drugs), and we will start to understand the basis of biological variation between cells. We aim to develop the computational methods needed to understand this wealth of data and to help guide experimentation. Our emphasis lies on project-oriented collaborations with experimenters. We work on the methods in statistical computing, integrative bioinformatics and mathematical modelling to turn these data into biology.