Similar protocols

Pipeline publication

[…] 83 of the 2,379 operons have experimentally confirmed TFBSs (solid bars in black) in their regulatory regions. Twenty of these 583 operons and their corresponding TFBSs were removed since they did not have enough orthology. The remaining 563 promoter sequences, containing 2,048 binding sites, were used to evaluate the performance of MP3. Besides, we downloaded Sigma 70 binding promoters of E. coli from the RegulonDB and conducted analysis to see the correlation between orthology and Sigma 70 binding in E. coli.Fig. 2, Performance evaluation. To conduct performance comparison, we applied six de novo motif finding tools previously mentioned, i.e., Biprospector, CONSENSUS, MDscan, MEME, CUBIC, BOBRO and a phylogenetic footprinting pipeline MicroFootprinter [–, , , , ] on the same genome and compared with MP3. We followed Tompa’s method [] and assessed the predictions both at nucleotide level and at the binding site level. Specifically, we calculated the sensitivity (nSN), positive prediction value (nPPV), specificity (nSP), performance coefficient (nPC) and correlation coefficient (nCC) at nucleotide level, and calculated the sensitivity (sSN), positive prediction value (sPPV), and average site performance (sASP) at site level. In addition, we added the widely used F-score (sFS) at site level for better evaluation. The calculation details for these measures can be seen in Additional file : Method S3. We followed Tompa’s criterion to indicate that a predicted site overlaps a known TFBS if the […]

Pipeline specifications

Software tools BoBro, MicroFootPrinter, DMINDA