Methods and strategies for construction of a phylogeny-adaptive hormone response element consensus model

Sex steroid hormones receptors bind to regions of the DNA called hormone response elements (HREs), in order to facilitate the regulation of gene expression. While the biological, functional and molecular basis of this interaction between the response elements and their corresponding transcription factor is not fully understood, the sequences of these HREs are known to be conserved for certain nucleotides. Machine learning processes have enabled researchers to predict and identify HREs in a quick and efficient manner. We had previously constructed a statistical model for HRE prediction from approximately 700 experimentally validated HRE sequences. To see if we can improve the performance of our statistical model for HRE prediction, we propose a phylogeny-adaptive model for HRE detection which takes into account the phylogenetic relationship of the sequences. The results of our analysis show that the new model ensures minimal false positive predictions.