Structure-aided prediction of mammalian transcription factor complexes in conserved non-coding elements.

Area of Science:

• Genomics
• Computational Biology
• Structural Biology

Background:

• Understanding transcription factor (TF) complex binding is crucial for deciphering cis-regulatory element functions.
• Existing methods often overlook complex binding site arrangements and physical constraints.

Purpose of the Study:

• To develop a computational method for predicting DNA-binding preferences of TF complexes.
• To identify evolutionarily constrained binding sites of rigid TF complexes.
• To analyze TF complex structures and their DNA interactions.

Main Methods:

• Compared co-occurring motif spacings in conserved versus unconserved human genome regions.
• Utilized structural data to assess TF complex physical plausibility and overlapping motif arrangements.
• Generated and analyzed 3D models of predicted TF complexes bound to DNA.

Main Results:

• Predicted 422 physically realistic TF complex motifs with an 18% false discovery rate.
• Identified that 77% of predicted complexes involve sequence overlap between binding sites.
• Discovered three distinct TF cooperativity mechanisms: direct, indirect, and 'through-DNA' interactions.

Conclusions:

• The novel computational method successfully predicts TF complexes and their binding sites, including those with overlapping sequences.
• Structural modeling revealed diverse mechanisms of TF cooperativity, with significant instances of synergistic binding.
• The findings provide a valuable resource for understanding gene regulation and TF-DNA interactions.