A novel computation framework for exploring densely mapped genetic markers, global gene expression and multiple physiological and
pathological phenotypes will be described in this talk. Many studies have shown that gene expression variation is heritable. The eQTL approach expands the traditional genetic study on the identification of the gene or genes directly responsible for a phenotype variation by treating the expression of a gene as a quantitative trait. Taking one step further, we ask how the genetic variation may affect the co-expression pattern of a pair of genes. Our system is based on the newly introduced concept of liquid association (LA). LA describes how variation in the pattern of
association between a pair of variables, including its sign and strength, is mediated by a third variable from the background. LA is introduced because despite the many successful applications of similarity based analysis on microarray data, there also exist numerous cases where the functional association is known from the literature (confirmed by experiments) but the statistical correlation from the expression data is practically zero. In addition to noises in the microarray data, a deeper
reason may be the biological complexity of the cellular system and the nature of its components such as multiple functions of a protein, varying
cellular oxidization-reduction states, fluctuating hormone levels or other cellular signals and the like.
This talk is based on work with Wei Sun, Robert Yuan and other members in the Bio-data refining group (http://kiefer.stat.ucla.edu/lap).
Copyright. All Rights Reserved.