Wednesday April 21st
University College London:
Malet Place Engineering Building, Room 1.02, UCL, 6pm
(Entrance via the Engineering Front Building where the cafe icon is)
Andrew Martin – SAPTF: Sequence Analysis Plugin Tool Framework
Sequence Analysis Plugin Tool Framework (SAPTF) is a simple application framework for developing web sites which perform analysis of one or more protein or DNA sequences. SAPTF uses a simple XML configuration file and can easily be integrated with more complex web site layouts, with all formatting being controlled by CSS. APTF uses AJAX to provide automatic handling of long-running programs which might make the user believe the web-server is not working, or which could cause the web-server to time out. The system is designed to call external programs (plugin tools) which perform the actual sequence analysis. These can generate HTML or simple plain text. SAPTF can pass sequence data to the plugin tools on the command line or in files and is able to manipulate the sequence data (for example, writing it in FASTA format) before calling the tool.
Alison Cuff – The CATH database: Structural Diversity and the Question of the Fold Continuum
CATH is a hierarchical protein structure classification database. This talk describes this classification system, and explores whether the structural divergence in some superfamilies and overlap between fold groups is posing a serious challenge to the concept of a hierarchical classification scheme.
John Pinney – GLASS: Gene LAyout by Semantic Similarity
Modern techniques of information visualization have a lot to offer the post-genomic biologist. As the ‘omic’ technologies continue to mature, resources to organise and display tens of thousands of data points will be required in order to present the results of many different types of genomic analysis in a coherent manner. Despite several advances in visualization techniques for bioinformatic data, there is still a notable lack of effective visualization methods for genome-scale data that incorporate existing biological knowledge in any meaningful way. At best, genomic data is presented as coloured nodes in a dense protein-protein interaction network, from which patterns are difficult to deduce. Better presentation methods for this type of data could help biologists towards a more intuitive understanding of the results of evolutionary analyses and high-throughput experiments.
In this talk I will present GLASS (Gene LAyout by Semantic Similarity), a new methodology for data visualization in a whole-genome context that produces attractive and informative graphical summaries of genome-scale data sets. GLASS organises the genome using the concept of the semantic similarity between genes, producing 2-D genome layouts in the form of Voronoi Treemaps. The maps created by GLASS can be coloured in various ways to visualize both discrete and continuous data attached to individual genes. Since genes with similar properties are grouped together, any biologically relevant patterns in the data are easily discernable and can be related in a natural way to the results of statistical analyses.
College Arms, Store Street, WC1E 7DH. We have booked the function room.