Predicting functionality of protein-DNA interactions by integrating diverse evidence

Area of Science:

• Molecular Biology
• Genetics
• Bioinformatics

Background:

• Chromatin immunoprecipitation (ChIP-chip) identifies protein-DNA interactions in vivo.
• ChIP-chip data alone is insufficient to determine the biological relevance of binding events.
• Regulatory interactions may be missed under untested conditions.

Purpose of the Study:

• To develop a probabilistic framework for identifying functional DNA binding events.
• To integrate ChIP-chip data with motif binding sites, nucleosome occupancy, and mRNA expression data.
• To determine the regulatory implications of transcription factor (TF) binding under various conditions.

Main Methods:

• Integrated ChIP-chip data with motif binding sites, nucleosome occupancy, and mRNA expression datasets.
• Employed a probabilistic framework to distinguish functional from non-functional DNA binding events.
• Analyzed condition-specific protein-DNA binding in budding yeast.

Main Results:

• Estimated that 50% of condition-specific protein-DNA binding in budding yeast is functional.
• Identified that binding functionality is highly condition-specific.
• Found that binding functionality is dependent on the presence of specific cofactors.

Conclusions:

• The probabilistic framework effectively identifies functional DNA binding events.
• A significant portion of observed protein-DNA binding is not functionally relevant.
• Understanding cofactor-dependent and condition-specific binding is key to unraveling transcriptional regulation.