Component 2 - Databases and Analytical Tools for Genomics and Proteomics
Zlatko Trajanoski Institute for Genomics and Bioinformatics, Graz University of Technology
Our goal in this component is to provide bioinformatics services and to develop databases and analytical tools for the exploration of
genomics, transcriptomics and proteomics data. We aim to maintain and continuously improve the environment we have developed in
the first GEN-AU period and meet new challenges.

Component 3 - Function Prediction with Biomolecular Sequence Analysis and Mass Spectrometry Data Interpretation
Alexander Stark Institute of Molecular Pathology
Component 3 will develop novel methods for the discovery of context-specific regulatory motifs and for the
prediction of the regulatory targets of these motifs and the corresponding factors. This is an essential step towards the
understanding and modeling of gene expression based on the regulatory genome sequence and as such is an important and integral
part of the BIN III goal of modeling biological networks. This component will also integrate tightly with the effort to understand
and model transcriptional regulatory networks in adipocytes (Component 2, Trajanoski), to which it will contribute motif-based
target predictions for the dynamic modeling of gene expression. Component 3 is timely with respect to data generation by ongoing
international projects including ENCODE and modENCODE. The methodology that will be developed will help to study general trends
and to develop rules concerning transcription factor binding and chromatin marks and their role in gene expression. The integration
of novel motifs and tools into a nucleotide annotation system will render them easily accessible and immediately useful for
experimental biologists. The combined interpretation of transcription factor binding sites, evolutionary conservation and novel
epigenetic data (DNA methylation, nucleosome positioning or histone-modifications) will provide complementary information and
therefore improve computational predictions and our understanding of any individual aspect.

Component 4 - Modeling and Inferring the Dynamics of Sequence Evolution
Arndt von Haeseler Center for Integrative Bioinformatics, Max F. Perutz Laboratories
Component 4 will develop a biologically realistic model of the insertion deletion process in biological sequences on the basis that evolution is influenced by a constraint-network. Modeling evolution under constraints is still in its infancy. To be able to model evolution more realistically there is an especially urgent need for methods able to infer complex dependencies from a multiple sequence alignment and to simultaneously take the tree structure into account. As well as theoretical models that are immediately applicable to current biological data, this component will also provide real implementations in the form of free software accessible for those who need them and expert advice regarding relevant methodologies. The ability to provide new models and methods upon demand and the expertise to apply these in a practical way represents an important resource for fundamental research. The project will moreover provide theoretical foundations for the fundamental processes in sequence evolution. The methods developed will also provide a first and very simplified view of how phenotype (sequence structure) and genotype (DNA or amino acid) influence each other during evolution. The methods and theoretical tools developed will moreover not only be applicable to study sequence evolution. They will also provide the building blocks for the modeling of the evolution of more complicated networks. The sequences will simply serve as one example to study the evolution of complex processes, because many and very reliable sequence data are available to work with.

Component 5 - RNA related Bioinformatics Tools
Ivo Hofacker Institute for Theoretical Chemistry, University of Vienna
Component 5 will contribute at the level of two-dimensional RNA structure by further developing methods to study the kinetics of RNA folding. This component will focus in particular on the problem of modeling the kinetics of inter-molecular duplex formation. In recent years RNA bioinformatics has been transformed from a niche field into one of the most active research areas. Tasks such as RNA secondary structure prediction, identification of cis-acting structures, detection of noncoding RNA genes and their regulatory targets have become highly relevant for many molecular biologists working on a wide range of biological problems. The expertise required to carry out such analyses is, however, still not widespread. Component 5 will therefore not only improve the quality and usability of tools for RNA structure analysis, but will also serve as a contact point for researchers from all other GEN-AU projects with respect to RNA related problems. In particular the component will focus on modeling complex formation between two RNAs, which should allow us to better understand the mechanism of translational control through non-coding RNAs such as microRNAs.

Component 6 - Computational Structural Genomics
Kristina Djinovic-Carugo Department of Structural and Computational Biology
Component 6 will develop methods for the comparison and classification of three-dimensional RNA structures. Although a number of techniques for the comparison of two (or more) three-dimensional protein structures have been designed during the last decades and have found many different applications, these methods are much less effective when the objects that must be compared are RNA molecules. Little is offered concerning RNA three-dimensional structures, despite the fact that several hundred RNA structures have been deposited into the Protein Data Bank. It is thus becoming urgent to develop tools able to handle this type of data in the manner that protein data are routinely handled. Given the available expertise, Component 6 is essential since it is the only one that can provide services in structural bioinformatics, ranging from structure prediction (secondary structure prediction, modeling, computational docking, etc.) to structure analysis (database mining, structure comparison, etc.). Its scientific activities perfectly fit these services, since both improvements in protein crystallization and RNA structure classifications will prove useful to the scientific community. The development of retrieval and classification tools for RNA three-dimensional structures is extremely well-timed. It mirrors the requirements faced by protein scientists 15-20 years ago and can be solved by adopting the same strategy based on locally developed databases and software (e.g. the very popular DALI server). The prospects for success are very great based both on external collaborations (University of Padova and JSI-Ljubljana) and the excellent interface between Components 1, 2 and 6. The final web-server is expected to play a prominent role in the structural biology of nucleic acids.

Component 7 - Data Mining in Proteomics
Christian Baumgartner University for Health Sciences, Medical Informatics and Technology
Component 7 will model correlation networks for contextualizing minimal marker sets and evaluating molecular interactions between conventional clinical chemistry measurements and omics data. Correlation networks will be constructed by leveraging pairwise analyte associations (i.e. Pearson, Spearman, Pearson-Partial or Spearman-Partial correlations) across and within compartments and disease states. A key goal is the construction of correlation networks for identifying interaction network master nodes (“hubs”) in-between, as well as across omics levels (e.g. proteomic ? metabolic data) by analyzing the systems response to different stimuli involved in abnormal regulatory and functional mechanisms in disease. The major goal in the third GEN-AU phase is the establishment of an integrative data mining, modeling and contextualization platform for advancing biomarker discovery and shotgun proteomics applications in a focused clinical context, thus clearly addressing the common goal of GEN-AU of improving prognostics, diagnostics and medical treatment of disease. This branch of research is an essential element in biomarker research and validation and translational medicine, transforming basic biomedical research to clinical practice.

Component 8 - SNAP (Systemic Network Analysis Platform)
Jacques Colinge Research Center for Molecular Medicine of the Austrian Academy of Sciences (CeMM)
Component 8 will develop SNAP (Systemic Network Analysis Platform), comprising innovative resources and methods for the integration and analysis of new drug-gene, drug-protein and protein-protein interaction data. Its generic design will make it applicable to a wide range of projects, in particular through a web front-end and open source release. The complementary developments described above will be integrated into a single web service hosted at the Graz University of Technology and will be made available to the biological community. Component 8 will provide a generic and in-depth platform to support research projects generating interaction data using any technology (proteomics pulldowns, expression data, co-evolution, etc.). GEN-AU projects APP III and PLACEBO will rely on BIN III Component 8 to perform the analysis of their interaction data and we strongly believe that SNAP will also benefit other research groups in Austria and abroad dealing with interaction data. In this sense, Component 8 is an essential component of BIN III.
The integration of established and useful network analysis methods, combined with the proposed innovative new directions in one platform will provide a powerful means of systematically extracting knowledge from interaction data and integrating various types of data in a network context. Such a platform will contribute significantly to interaction network analysis research and application.