Simulating the evolution of transcriptional regulatory motifs

Abstract

Transcription factors are proteins that bind DNA for the purpose of regulating the expression of specific genes. The process by which transcription factors identify their binding sites is poorly understood. It is possible to infer the nature of recognition strategies used by transcription factors through an examination of the information content of their associated binding motifs. Unfortunately, experimentally verified binding motifs are often incomplete and numerical methods of examination frequently suffer from under-sampling artifacts. We attempt to better understand the feasibility of specific recognition strategies by simulating the co-evolution of a transcription factor and its associated binding motif. We explore the hypothesis of positional independence and conclude that it is unlikely for an organism to rely heavily on correlations between positions in a binding motif as its primary means of binding site recognition.