Mapping Operator Sequence to Transcription Factor Binding Energy <i>In Vivo</i>

Despite the central importance of transcriptional regulation in systems biology, it has proven difficult to determine the precise regulatory mechanisms of individual genes, let alone entire gene networks. The advent of readily-available DNA sequencing has opened up numerous ways to interrogate promoters and transcription factors. Yet, it is still challenging to integrate the various aspects of transcriptional regulation into a cohesive understanding of a given promoter or transcription factor. One promising avenue for elucidating multiple aspects of transcriptional regulation with a single assay is through Sort-Seq, a high-throughput in vivo assay which associates expression measurements with the sequences of millions of promoter mutants. Here, we show how Sort-Seq can be used to accurately map binding site sequence to binding energy and predict the binding energies of operator mutants to within 1 k_BT of their measured values. We further explore how such a sequence-energy mapping can be used to address key scientific challenges, such as designing specific induction responses, analyzing the probability that a mutation will affect the fitness of a promoter, or determining how regulatory context affects sequence specificity.