Detecting Circadian Genes from Microarray

C. LAYANA; DIAMBRA, L.

Montevideo

Congreso; ISCB LATIN AMERICA 2010; 2010

International Society for Computational Biology

Microarray experiments allows estimation of relative expression of thousands of genes at each time point, and they are widely used for monitoring gene activities in a cell during biological processes. However, the data produced in these experiments carries a high degree of stochastic variation, obscuring the periodic pattern. Identification of circadian expression pattern in time series data is challenging. Because the measured data are often non-ideal, and efficient algorithms are needed to extract as much information as possible.We analized the results of microarrays experiments of circadian cycle in Cyanobacterium Synechocystis [1] by using autoregressive modeling of the expression dynamics and define a model-based distance between two profiles. Our approach is based on the Maximum Entropy Principle in the Information Theory framework [2,3]. This tool allow to us detect gene with circadian expression and quantify how different are the expression dynamic of two genes.In order to detect the genes of the core clock, we selected the genes that satisfy three condition: I) have circadian expression, II) have similar dynamics in both biological replicated, and III) the phase shift is small. Figure 1 illustrates the data allow a further analysis: the dynamic and phase comparison between replicates in order to select genes candidates to be clock genes. Dynamic distance vs Phase shift between the oscillatory genes of both replicates. Dots green are genes which dynamics is similar and phase is synchronized in both replicates. Between these 88 genes we identify two genes, black dots (slr1942 and slr0756) which are related to the clock machinery. Another two genes were did not identified with these criteria.In conclusion, we propose that our procedure is a promising statistical tool for finding oscillatory expressed genes of any period in a microarray data set.