Knowledge Management Enviroments for High Throughput Biology

Abstract

With the growing complexity and scale of data sets in computational biology and chemoin- formatics, there is a need for novel knowledge processing tools and platforms. This thesis describes a newly developed knowledge processing platform that is different in its emphasis on architecture, flexibility, builtin facilities for datamining and easy cross platform usage. There exist thousands of bioinformatics and chemoinformatics databases, that are stored in many different forms with different access methods, this is a reflection of the range of data structures that make up complex biological and chemical data. Starting from a theoretical ba- sis, FCA (Formal Concept Analysis) an applied branch of lattice theory, is used in this thesis to develop a file system that automatically structures itself by it’s contents. The procedure of extracting concepts from data sets is examined. The system also finds appropriate labels for the discovered concepts by extracting data from ontological databases. A novel method for scaling non-binary data for use with the system is developed. Finally the future of integrative systems biology is discussed in the context of efficiently closed causal systems.