Main Research Highlight
Although interactions between transcriptional factors (TFs) and genes through low-affinity promoters have been largely ignored and their contribution in transcriptional regulation disregarded, they are abundant in vivo. These bindings between TFs and low-affinity promoters form a notable fraction of the interactions between TFs and DNA and are promoter sequence dependent.
Strongest Evidence Supporting this Research Highlight
The strongest evidence that supports this highlight is the fact that through analyzing data from the current Chromatin Immunoprecipitation (ChIP) experiments quantitatively, the study reveals widespread low affinity transcriptional interactions in a genome. Unlike previous analysis methods the ChIP data was interpreted quantitatively, the focus was therefore not only on a few dozens high specificity hits for each TF but covered the entire specificity range of the TFs (Tanay, 2006).
Why the quantitative approach is the strongest evidence
The previous working hypothesis that TFs either bind perfectly to a sequence motif or do not bind at all meant that the analysis of ChIP data made use of a digital model such that all the data was transformed into binary TF-gene interactions. This openly ignores the contribution of low-affinity TF-gene interactions that do not give perfect binary ChIP measurements. The quantitative approach employs an analogue model that analyses the ChIP data as it is without any transformations thus increasing the possibility of detecting low-affinity TF-gene interactions.
The ChIP data from mapping yeast genome transcriptional network indicated information content in nonspecific binding profiles. The occurrence of this may be due to TF-DNA interactions which are functionally organized although they are not indicating any highly specific interactions. Such information content may also be as a result of experimental or normalization artifacts thus nullifying the research highlight of the paper. To strengthen the highlight, ChIP binding ratios were compared to the sequence-based predictions of TF affinities. The predicstion is usually done using the Position Weight Matrices (PWM). The ChlP data and PWM predictions showed high correlation over a wide dynamic range. This indicates high dependence between the ChlP values and the sequence hence greatly supporting the fact that TF-low affinity promoters binding is promoter sequence dependent and that the quantitative approach is advantageous. It can be seen from this that ChlP profiles can able to provide information about a wide dynamic range of specificities.
The possibility of performing regression of a PWM model to an entire ChIP binding profile by use of PREGO algorithm further strengthens the quantitative approach in transcriptional analysis. Since this algorithm exploits information from the full spectrum of binding energies, it is possible to detect PWMs that could not be detected previously from the ChlP data using the digital model. These PWMs are for the low-affinity transcriptional interactions that the digital model ignored. The PREGO algorithm therefore ensures that quantification of the ChlP values is possible and the magnitude of the low-specificity TF-DNA binding can be determined.
Another benefit of using PREGO analysis that helps strengthen the quantitative approach is the ability to characterize TFs' binding preferences even when ChIP analysis generates few or no high-specificity hits for a certain TF. In such circumstances, the previous analysis methods would not detect any PWM.