Search tool for motif/ChIP-based binding site clusters. (A) An interface to search for motif/ChIP-based binding site clusters. There are three main panels: Search options, Motif collection and ChIP data collection. Search options: here, the user may select the types of profiles (‘Sorting criteria’) to use in the search and also the search region, which is either the entire genome or a list of genomic loci. More advanced search options are also provided (e.g. the number of top results). Motifs collection and ChIP data collection: here, the user may select the specific motifs and ChIP profiles to use in the search. (B) Sample output of the search tool. A genome-wide search was conducted with five TFs involved in anterior patterning (BCD, KR, HB, TLL and HKB), retrieving segments (of 500 bp each) where a significant cluster of binding sites for one or more of these TFs was found. (Only the top two reported segments are shown in the figure.) The second column reports the location of the segment, along with a link to the Genome Surveyor view of this locus. It also provides the names of known CRMs that overlap the segment (e.g. ‘eve_stripe1′ for segment 1). The third and fourth columns show the strength (z-score relative to the genome-wide average) of binding site clustering in the segment, based on single species [‘Stubb(SS)’] and multi-species [‘Stubb(MS)’] analysis respectively. The fifth column reports the scores of each selected motif, separately for single species (‘SS’) and multi-species (‘MS’) analysis. The last column provides information about the nearest neighboring genes of the segment.

Genome Surveyor 2.0 input/output. (A) General Scheme of Genome Surveyor 2.0. Users may investigate the region of interest (defined through GBrowse) for potential cis-regulatory activity using a variety of tracks. To find the targets of specific motifs, users may activate the corresponding single species, multi-species and/or ChIP profiles. They may search for CRMs that are similar to a set of previously validated CRMs, using the ‘Supervised CRM discovery’ tracks. The location of strong binding sites in a region can be visualized with the ‘Binding sites above threshold’ track. Known CRMs are displayed through ‘REDfly CRMs’ track. (Green, red and blue boxes show inputs, built-in tracks and tracks that are computed on-the-fly, respectively.) (B) An instance of Genome Surveyor output. Shown is the 20 k region surrounding the eve gene. Thirteen different tracks are shown: (numbered from 1 to 13 top to bottom) ¹REDfly CRMs and ²Predicted genes. ^3,4Motif profiles for Bicoid (Bcd_SOLEXA) in D.mel and D.pse as two separate tracks (selected from the ‘Stubb: individual motifs, individual species’ plugin). ^5,6Multi-species profiles for Bicoid (Bcd_SOLEXA) and Kruppel (Kr_NAR) motifs (selected from the ‘Stubb: individual motifs, multi-species avg.’ plugin). ^7,8Average of motif profiles for Bicoid (Bcd_SOLEXA) and Kruppel (Kr_NAR) motifs in D.mel and D.pse as two separated tracks (selected from the ‘Stubb: multi motif avg., individual species’ plugin). ⁹Average of multi-species profiles for the Bcd_SOLEXA and Kr_NAR motifs (selected from the ‘Stubb: multi-motif, multi-species avg.’ plugin). ¹⁰Motif profile for user provided PWM; here, the Bicoid (bcd) motif from the JASPAR database (selected from the ‘Stubb: your motifs’ plugin). ¹¹ChIP binding profile for Bicoid (selected from the ‘ChIP Tracks’ plugin). ¹²The similarity profile to a set of known CRMs active in blastoderm-stage development (selected from the ‘Supervised CRM discovery’ plugin). ¹³Individual high scoring sequence motif matches (blue codes for the highest score) for Bcd_SOLEXA (selected from the ‘Binding sites above threshold’).

Example of motif composition analysis in known CRMs. Shown is the 21-kb region surrounding the hairy gene. Eight different tracks are shown: the ‘Stubb individual species’ profiles for seven TFs (CAD, KR, BCD, GT, HB, KNI, TLL) involved in initial stages of anterior–posterior (A/P) patterning in the embryo, and the ‘Stubb multi-motif avg.’ profile over TLL, KR, BCD and GT. Blue shaded regions represent two known A/P related CRMs, h_stripe1 and h_stripe0, that are both known to be BCD-activated CRMs (21). The presence (high peaks in score profile) of this known activator, as well as specific transcriptional repressors (TLL and KR for h_stripe1 and GT for h_stripe0), is consistent with the CRM expression patterns. The bottom profile (red) illustrates how combining Stubb profiles from multiple motifs might help in CRM discovery.