Integrative Analysis of Transcriptional Regulation in E. coli

From the intensity files, macs2 software (7 (link)) and BEDOPS (8 ) commands were run to obtain enrichment regions. These sequences were processed with matrix scan from RSAT tools (9 (link)) to obtain putative TFBSs. A HT TFBS and a RegulonDB TFBS were considered the same if they overlapped in >50% of their length. Conditions or phenotypes of RNA-seq experiments were identified to generate a condition contrast. The two conditions in RNA-seq datasets were compared with a two-sample t-test and the resulting p-values were adjusted for the false discovery rate (FDR). The log₂ fold change across conditions and the FDR P-values were used to construct a volcano plot (10 (link)) highlighting the differentially expressed genes (FDR < 5%). To identify RIs, the TFBSs were mapped, when possible, to the regulatory regions of Escherichia coli genes using an ad hoc python script. A regulatory region per gene per TF was defined either by the distance to the farthest known TFBS (with strong evidence) of such TF or as the interval between −400 and +100 base-pairs with respect to transcription initiation. The processing of files is shown in Supplementary Figure S1.