”Trigonometry is a sine of the times” — unknown



Circadian rhythms are ubiquitous and exhibit exquisite precision giving rise to balance and homeostasis. Permissive gating plays a pervasive role in temporal compartmentalization of key processes including regulation of growth and development, coordination of metabolism, flowering time, stress responses, immunity and plant pathogen interactions resulting in greater fitness and adaptability. Transcriptional networks driving patterns of gene expression represent important form of plant signaling. They give rise to versatility and flexibility allowing cells to determine their actions in response to external environment and internal signals. The aim of the investigation is to delineate higher order structure of the circadian module that coordinates expression over the course of phenotypic change, through in silico learning. This is done by recourse to reverse engineering of expression data obtained from the model plant Arabidopsis thaliana. Cis and trans acting elements together regulate initiation of transcription. DNA governs complex formation as nucleotide composition of the binding site affects coregulator recruitment. So far, protein centered approaches have aimed to delineate binding behaviour of one protein at a time. Here a sequence-bait centered method is used for high throughput identification of regulatory motifs. The methodology involves several predictive modeling approaches including Naive Bayes, simple decision tree methods, random forest and stepwise linear regression resulting in efficient interpretable predictions. They allow for delineation of prominent sequence features through selection of variables with the aim being prediction of phase, that is, gene expression, relying on a set of features or motifs. Initial results have generated testable hypotheses that are yet to be confirmed. Sets of regulatory loci working in synergy have been defined and will be used to infer architecture of the plant circadian clock.

1 2 3