TFBSshape provides DNA shape features for transcription factor binding sites (TFBSs) that in addtion to sequence features, usually in the form of position weight matrices (PWMs), characterize DNA binding specificities of transcription factors (TFs) from 23 different species. The DNA shape features were predicted using our recently published high-throughput method (1) and visualized in the form of heat maps. Users can download these DNA shape feature data and perform further analysis, for example, to identify which DNA shape feature(s) at which position(s) in the target DNA contribute to the DNA binding specificity of a TF. Users can also compare DNA shape profiles, both qualitatively and quantitatively, between two TF datasets from the database, or between a set of uploaded sequences and a TF from the database. TFBSshape collects data from the open-access motif databases JASPAR (2) and UniPROBE (3). However, TFBSshape is fundamentally different from the two databases in that it provides DNA shape profiles for TFBSs in addition to sequence information. The JASPAR database provides TFBSs for a significant portion of its TF collection. For these TFs, TFBSshape uses our high-throuput method to predict DNA shape features for the TFBSs. On the other hand, the UniPROBE provides only protein binding microarray (PBM) probes and PWMs derived from the PBM experiments for the TFs, instead of specific TFBSs. In this case, TFBSshape first uses FIMO (4) to scan the PBM probes for TFBSs given the probe sequences and UniPROBE PWMs as input. DNA shape feature are then predicted for these TFBSs found by FIMO. Since DNA shape has been shown to play an important role in protein-DNA recognition (5), providing DNA shape feature data for TFBSs derived from motif databases will help to advance our understanding of the mechanisms underlying TF-DNA interactions.




nucleic acid sequence transcription transcription factors and regulatory sites methylated dna immunoprecipitation epigenetics genetic variation chip-seq

More to explore:


Need help integrating and/or managing biomedical data?