Rewiring the specificity of extra-cytoplasmic function sigma factors

Bacterial genomes are being sequenced at an exponentially increasing rate, but our inability to decipher their transcriptional wiring limits our ability to derive new biology from these sequences. De novo determination of regulatory interactions requires accurate prediction of regulators DNA binding and precise determination of biologically significant binding sites. Here, we address these challenges by solving the DNA-specificity code of extra-cytoplasmic function sigma factors (ECF {sigma}s), a major family of bacterial regulators, and determining their regulons. We generated an aligned collection of ECF {sigma}s and their promoters by leveraging the auto-regulatory nature of ECF {sigma}s as a means of promoter discovery and analyzed it to identify and characterize the conserved amino acid - nucleotide interactions that determine promoter specificity. This enabled de novo prediction of ECF {sigma} specificity, which we combined with a statistically rigorous phylogenetic foot-printing pipeline based on precomputed orthologs to predict the direct targets of [~]67% of ECF {sigma}s. This global survey indicated that ECF {sigma}s play varied roles: some are global regulators controlling many genes throughout the genome that are important under many conditions, while others are local regulators, controlling few closely linked genes in response to specific stimuli. This analysis reveals important organizing principles of bacterial gene regulation and presents a conceptual and computational framework for deciphering gene regulatory networks.