Figure 1.
The used approach to genome-wide selection of rSNPs.
Computational analysis was applied to identify the SNPs in the most likely regulatory regions of the human genome and predict rSNPs for experimental verification.
Figure 2.
The enrichment of functionally-associated Somim, Sclinic, Sgwas samples, and Sr sample with putative rSNPs as a function of cut-off number of overlapping TF binding loci (i) for defining OTFRs.
500 bootstrap iterations were performed for each point. The resulting standard deviations and confidence intervals are shown by error bars and colour-filled areas,respectively. Sclinic, Somim, and Sgwas consist of SNPs associated with phenotypic manifestations and extracted from dbSNP NCBI Clinical/LSDB Submissions Resources, OMIM catalog, and NHGRI GWAS catalog, respectively. Sample Sr of random SNPs was created without applying any phenotypic preferences and used as control. Genomic region chr6:29,909,708-31,325,212 was excluded from the analysis of Sclinic sample, since it caused enrichment overestimation at lower i (a large number of SNPs concentrate in this region due to its extensive use in genotyping assays).
Figure 3.The.
enrichment of Sgwas sample and its high-confidence derivatives, SpV, SOR and Sint, as well as Sr sample with putative rSNPs as a function of cut-off number of overlapping TF binding loci (i) for defining OTFRs.
500 bootstrap iterations were performed for each point. The resulting standard deviations and confidence intervals are shown by error bars and colour-filled areas,respectively. The subsamples of Sgwas were generated with filtering of SNPs by P-value <1 e–7 (SpV), OR>3 and <0.3 (SOR), and by both criteria (Sint). Sgwas sample was extracted from NHGRI GWAS catalog.
Figure 4.
Localization of putative rSNPs within OTFR belonging to APC gene.
SNPs from the different parts of OTFR were taken in EMSA to study the effect of SNP location within OTFR on the protein binding.
Figure 5.
A variety of SNP effects on binding of the corresponding oligonucleotides to nuclear proteins from K562 cells.
rs79734816:C>T (A), rs2071002:A>C (B), and rs74393987:C>T (C) change the number and intensity of bands, while rs75996864:G>T (D) affect only band intensity, and 7961894:C>T (E) do not have any. Changes in the binding of allelic variants with the nuclear proteins are indicated by arrows.
Figure 6.
Cell line-dependent binding of nuclear proteins to oligonucleotide probes.
Binding of nuclear proteins isolated from HCT-116 (1, 2), HeLaS3 (3, 4), K562 (5, 6) and HepG2 (7, 8) human cell lines to the V1 (C- allele) and V2 (T- allele) oligonucleotides corresponding to allelic versions of rs79734816:C>T.
Table 1.
Summary of the experimental testing of putative rSNPs selected from Somim.
Table 2.
Summary of the experimental testing of putative rSNPs from different parts of OTFR belonging to APC gene.