Our programme of guest and selected oral and poster presentations comprised a diverse range of biologically and clinically relevant questions, computing approaches and applications . These sessions were driven also by a significant interest in discussing challenges and directions for enhancing the impact of bioinformatics in fundamental and translational biomedical research. These features were specially reflected in the presentations given by our guest speakers: Burkhard Rost (Technical University Munich, Germany), Thomas Lengauer (Max-Planck-Institut Informatik, Germany), Ioannis Xenarios (Swiss Institute of Bioinformatics, Switzerland), and Peter van der Spek (Erasmus Medical Center, The Netherlands).
Presentations and discussions addressed applications of significant relevance to the detection and treatment of cancers, infectious, neurodegenerative and cardiovascular diseases. The meeting also showcased progress in bioinformatic infrastructures and enabling technologies. Examples of the latter included next-generation sequencing tools and applications, complex information integration and annotation over the Web, and several disease- and organism-specific research resources.
Among such a diversity of approaches and applications, a crucial pattern can be discerned. Bioinformatics research is putting a great emphasis on answering "when", "what if" and "why" questions using different types of biological data. When are complex events at the molecular and systems levels taking place ? When are specific disease states likely to occur? Thomas Lengauer's talk on translational bioinformatics in treating HIV infections represented an instance in this direction. What if specific perturbations are induced ? Why are observed associations or predictions biologically meaningful ? Ioannis Xenarios's talk on gene and cellular regulatory network models was an example within this research question category.
Different presentations also highlighted opportunities for integrating multiples vistas of biological phenomena as dynamic webs of interactions. Examples of this included: "Visualizing genotype-phenotype relationships across cell cycle and evolutionary time scales" (Maria Secrier and Reinhard Schneider), "Prediction of a phosphorylation network in Arabidopsis thaliana" (Kris Laukens et al.) and "Interfering with the interaction network of adenyl cyclase virulence factors" (Therese Malliavin et al.). This and future contributions would not be possible without advanced capabilities in large-scale data analysis and interpretation, for example from next-generation sequencing experiments (Alejandro Sifrim et al.; Tim de Meyer et al.) and from other types of "omic" data (Yan Wu et al.; Alexey Stukalov et al.).
Harnessing the huge amounts of data routinely produced by high-throughput techniques remains a challenge, as demonstrated by some of the posters presented during the conference. While microarrays still represent a fair amount of the contributions in genomics, next-generation sequencing has begun its journey towards clinical applications. This is happening thanks to the development of affordable, robust and well-standardized pipelines available for service, as introduced in "Union makes strength: building baseline Tracks from 69 open access full human genomes" (Stephane Plaisance and Mark Veugelers). Having all this information available, efforts are being put in its "organisation": building pathways, modelling networks, analysing interactions, and prioritising candidates for further downstream validation. "Analyzing gene and protein expression variance in cellular pathways using high-throughput experimental data" (Enrico Glaab and Reinhard Schneider) and "Network analysis of differential expression for drug target prioritization" (Griet Laenen et al.) are examples of work in this direction. Given all the methods proposed, it is also worthwhile to compare how they perform, as demonstrated in "Critical assessment of candidate gene prioritization methods" (Daniela Bîrnigen et al.).
Another remarkable point is how bioinformatics is expanding beyond research environments to reach the clinicians directly: "Breaching the surface with HOPE" (Jules Kerssemakers) aims at abstracting the complexity behind resources dealing with protein interfaces, surfaces and assemblies, thus helping physicians to understand the structural effects of mutations. Furthermore, one of the three awarded posters, "Clinical data miner - an electronic data capture software framework that improves interrater agreement" (Arnaud Installé et al.), was specifically aimed at the routine clinical environment. The proposed method enhances expert robustness and facilitates interpretation in situations where "case report forms" are used.