Structural Motifs (StruMs) are a model of transcription factor binding specificity derived from DNA shape. Today we released an optimized version of the StruM Python package, that uses Cython to make the computations more efficient.
To install the package or learn more, visit the GitHub: Structural Motifs
Here is a sneak peak from the documentation:
StruM: Structural Motifs
The interaction between sequence specific transcription factors and DNA requires compatability at a complex interface of both electrostatics and 3D structure. These features depend on the sequence context of the motif, with dependence between adjacent nucleotides.
StruMs use estimates of DNA shape to model binding sites for sequence specific transcription factors. Here we provide tools to:
- Train models from known binding sites
- Use expectation maximization to find de novo motifs
- Provide utility to incorporate additional quantitative features, e.g. DNase hypersensitivity.
- Score matches to novel sequences
For full documentation, refer to the docs here (html) or here (pdf).