Furthermore, based on the assumption that two protein families usually interact in a similar way, with the same binding site, we predict all four LGI family members to use this interface to interact with different ADAM proteins, albeit with different affinity, in a time and space dependent manner.
By applying three different weight matrices for recognition of c-Myc binding sites, we predicted 94 putative c-Myc targets among the genes presented on Affymetrix platform GeneChip® Mouse Genome 430 2.0.
We refined this list by focusing on target genes whose expression was inversely correlated with the miRNA expression (i.e. potentially downregulated), and using a Targetscan context + score of < −0.30 for at least one binding site, we identified five predicted gene targets for miR-29b and four for miR-223 (Table 4).
To compare the selected sets of the putative binding sites, we computationally predicted all binding sites in the human genome using our affinity data (Additional file 5).
To test the specific regulation through the predicted binding site, we constructed a reporter vector consisting of the luciferase coding sequence followed by the 3′-UTR of Smad4.
To decide a relevant threshold for accepting a predicted binding site, we repeated the analysis in randomly selected regions with the respective PWM.
For each predicted binding site, we listed the nodal and tip events according to previously published parameters (Levin et al. 2013).
While the human tetherin cytoplasmic tail contains a predicted TRAF6 binding site, we found that this interaction was actually determined by the adjacent endocytic motif YDYCRV.
To evaluate whether SNPs occur in the experimentally derived TF binding sites, we linked the predicted TF binding sites with the binding sites in TRANSFAC using RefSeq literature references.
Furthermore, Pak3, although highly homologous to the other Group 1 Pak members, does not contain a clear LC8 binding site and we predict that its nuclear localization is not LC8 dependent.
