The DNA-binding landscape

GENE REGULATION

The DNA-binding landscape

Proteins or small molecules designed to bind certain DNA sequences and to regulate target genes have much promise in both basic and applied research. Aseem Ansari and colleagues at the University of Wisconsin, Madison, therefore tested the specificity of factor binding to DNA. They used custom DNA arrays displaying every possible 10-mer sequence—almost a half a million of them—to examine the binding profiles of engineered hairpin polyamides and protein transcription factors, as well as of several natural DNA-binding proteins, across sequence space. They quickly ran into a problem. The data were just too complex to understand intuitively. “We had this comprehensive binding dataset,” says Ansari, “and the first question was, how do you look at these data? If you use colors or tables or graphs to represent it, things begin to get very complicated very quickly.” They saw that protein transcription factors, in particular, bind fairly broadly across sequence space and that, although it is possible to extract a consensus binding motif, this often did not tell the whole story. What is more, the consequences of changing particular residues in a binding motif were not always simple or additive. So they had to find a new way of visualizing the data.

Nature Methods 7, 254 - 255 (2010)
doi:10.1038/nmeth0410-254a