Affiliations: 1) Pennsylvania State University; 2) Arkansas State University
Keywords: Genomic selection/prediction
The evolution of gene expression is critical for adaptation due to its role in shaping organismal function. However, predicting the response to different environmental conditions is complicated as variation in cis regulatory elements, including transcription factors, impact the potential of adaptation to novel environments. Incorporating information about the variance in the genetic code across multiple genotypes from different populations may improve the prediction of response to environmental fluctuations and give insights into the evolution of expression and plasticity. We use machine learning approaches to predict the impact of cis regulatory variation on stress responsive expression across Arabidopsis thaliana populations. The predicted variation is then tested for conservation across populations. Thus, we explore regulatory elements that are universally important for response to the stress, as well as novel within specific lineages. The impact of selection on the predicted sites is studied, identifying the relative importance of selection in evolving the regulation of expression under stress.