Biological research methodologies are changing rapidly. Not so long ago, biology was a largely descriptive and subjective science. The advent of high-throughput quantitative "-omic" technologies has ushered in a new era characterized by large-scale experimental approaches in biology that were not previously feasible. These technologies are able to provide information on many measured variables at once, in sharp contrast with earlier methods where one or at most a few variables could be simultaneously investigated. These new capabilities create an opportunity to design experiments in new ways. Rather than forming a very specific hypothesis and then designing an experiment to test that hypothesis, researchers are now able to create experimental designs capable of generating broad classes of hypotheses that are testable by analysis of the massive data sets that result. Such high-dimensional data sets are potentially rich in information regarding their underlying biological systems, but discovering this biologically-meaningful information is becoming a new science in its own right. In this talk I will describe pattern discovery and datamining techniques developed at Cira Discovery Sciences, and their application in discovering new markers for leukemia/lymphoma (cytomics) and for preterm labor in pregnancy (proteomics).
