(114b) A Gene-Ontology Driven Analysis of Complex Phenotypes

Complex phenotypes of cellular systems result from multilevel and multivariable cell regulation, and are thus difficult to deconvolute by reductive approaches alone. An example of a complex phenotype is the metabolite-stress response of differentiating (endospore-forming) bacteria such as bacilli and clostridia. In clostridia, product formation, stress response and pathogenesis are intimately related to and affected by differentiation, and thus a fast-track deconvolution of this complex interplay is of significant interest. Another example is stem-cell differentiation and maturation into a particular cell type, such as into megakaryocytes (the precursors of platelets) from ex vivo expanded human hematopoietic stem cells. Megakaryocytic differentiation and maturation are centered on programmed-cell-death (apoptosis; the death of megakaryocytes gives rise to platelets), and involves transcriptional regulation, signal transduction, and posttranslational activation of proteins. To deconvolute such complex phenotypes one would need complete genomic data at the transcriptional level (by DNA microarrays) and at the protein level, including posttranslational modification, etc, plus substantial computational tools. While DNA-microarays are finally reaching the point of providing complete (genomically speaking) and, at least, semi-quantitative data, this is not the case yet with protein level data. I will discuss how gene-ontology (GO) analysis of DNA microarray data coupled with other reliable low and high-throughput tools can be recruited to enhance our understanding of biological complexity.