Paver

Paver is a DECODON service for the analysis and visualization of large OMICs data sets, including data from transcriptome, proteome and / or metabolome analyses. A newly developed integrated visualization technique aggregates complete sets of expression data with available biological knowledge from a variety of gene functional classification systems on a minimum of screen space. This makes it possible to visualize expression sets as a whole without any kind of pre-filtering.

Paver combines expression data and functional classification to build a visual display of all detected mRNAs or proteins of the biological system under investigation. In contrast to other approaches Paver can integrate the whole set of experimental data with a maximum of biological knowledge already available.
Gene functional classification systems organize genes / proteins according to their function in (quasi-)hierarchical systems. Paver supports:

• Gene Ontology
• KEGG Brite
• Riley scheme derivatives (e.g. TIGR classifications, Genolist classifications)
• Clusters of Orthologous Groups (COG)
• Open Biologicy Ontologies (OBO)
• gene regulatory data
• custom classification schemes.

Paver finds and clearly highlights genes and proteins that behave similarly; grouping by function makes them available for further interpretation and analysis. This allows bio-scientists to find interesting functional coherencies within biological systems and to discover new patterns never seen before in their expression data.

Voronoi Treemaps - Paver's technology for data visualization

Treemaps were established by Ben Shneiderman from Maryland in the mid 90's to display hierarchical data using nested (rectangular) areas. Each branch of a classification hierarchy is represented by a rectangular area. Subbranches tile them into rectangular subshapes and so on. This efficiently uses available screen space but the use of rectangular shapes makes perception of hierarchical structure difficult. Voronoi treemaps replace the rectangular areas by convex Voronoi shapes (Balzer & Deussen).

These shapes provide improved visual perception of hierarchy levels and related entities. When applied to a regulatory classification scheme from bacteria, the layout of a section of a treemap looks as follows:

Parts of a map - genes are represented by the smallest subunit, regulators are the largest.

In a similar manner layouts based on Voronoi treemaps can be applied to any imaginable functional classification system.

Changes in expression levels for genes or proteins are visualized using a color gradient. Multivariate data (expression change combined with uncertainty and amount of biomolecules) may be represented using gray shading (i.e. saturation - see figure) or variation of brightness.

Color scales and grey shading – used to visualize expression changes and uncertainty.

The Paver service

Expression data from any field of OMICS technologies are mapped to gene functional classification systems and analyzed. DECODON's service team guides you through the whole analysis process. New (customized) classification schemes of interest can be implemented if necessary. You'll be provided with sets of visualizations, statistical data on significantly influenced functional groups.

Development

Paver's underlying engine has been developed in a collaborative effort of DECODON GmbH and the group of Dr. Jörg Bernhardt at the Ernst-Moritz-Arndt-University of Greifswald bringing together expertise in software development, bioinformatics and OMICS centric biology.

Until now Paver and its underlying technology have been applied in molecular medicine (human cell lines, treatment of blood products), pathogenomics (host pathogen interaction, response of human cell lines to infection) as well as microbial physiology (B.subtilis, S.aureus strains, Pseudomonas strains) and MetaOMICS related studies.

Learn more

Just call us at +49 (0) 3834 515230 or send an email to

info@decodon.com

with the subject "Paver information request".

We are looking forward to hearing from you!

References

• Bernhardt J, Mehlan H, Schüler J, Hecker, M (2011). OntologyMaps. Awarded as Best Poster at VizBi (Visualization in Biology) 2011 Conference at the BROAD Institute in Boston, MA, USA.
• Otto A, Bernhardt J, Herbst FA, Siebourg J, Mäder U, Hecker M, Becher D (2010). System-wide temporal profiling of proteins in glucose starved Bacillus subtilis. Nat. Commun. 1:137.
• Schmidt F, Scharf SS, Hildebrandt P, Burian M, Bernhardt J, Dhople V, Kalinka J, Gutjahr M, Hammer E, Völker U (2010). Time resolved quantitative proteome profiling of host-pathogen interactions: The Staphylococcus aureus RN1HG response to human S9 cells after internalization. Proteomics. 2010 Aug;10(15):2801-11.
• Bernhardt J, Funke S, Hecker M, Siebourg J (2009). Visualizing gene expression data via Voronoi treemaps. Voronoi Diagrams in Science and Engineering, 2009. ISVD '09. 6th International Symposium on. 0: 233-241. ISBN: 978-0-7695-3781-8.