| Popis: |
Rosalind B Penney,1 Abbie Lundgreen,2 Aiwei Yao-Borengasser,3 Vineetha Koroth-Edavana,3 Suzanne Williams,3 Roger Wolff,2 Martha L Slattery,2 Susan Kadlubar3 1Department of Environmental and Occupational Health, University of Arkansas for Medical Sciences, Little Rock, AR, 2Department of Internal Medicine, University of Utah Health Sciences Center, Salt Lake City, UT, 3Division of Medical Genetics, University of Arkansas for Medical Sciences, Little Rock, AR, USA Abstract: There are a number of in silico programs that use algorithms and external web sources to predict the effect of single nucleotide polymorphisms (SNPs). While many of these programs have been shown to predict accurately the effect of SNPs in functional areas of the gene, such as 5΄ upstream or coding regions, empiric research may be warranted to confirm the functional consequences of SNPs that are predicted to have little to no effect. We compared predictions from FASTSNP (Function Analysis and Selection Tool for Single Nucleotide Polymorphism) and F-SNP (Functional Single Nucleotide Polymorphism) with experimentally derived genotype-phenotype correlations to determine the accuracy of these programs in predicting SNP functionality. We used normal colon tissue to evaluate 24 TagSNPs within six genes. Two of 16 SNPs that were predicted to have no functional effect in FASTSNP were significantly associated with gene expression. Only one of the eight SNPs that were predicted to have a low to high effect was significantly associated with gene expression. While the two in silico programs that were used were similar in their results for the SNPs predicted by FASTSNP to have no effect, of SNPs with scores from low to high, there were three that received an F-SNP score below what is considered functionally significant. In silico programs can fail to identify functional SNPs, supporting a continuing role for empiric analysis of SNP function. Laboratory analysis is necessary to identify causal SNPs accurately, establish biological plausibility of the effect, and ultimately inform cancer prevention strategies. Keywords: in silico prediction, colon, single nucleotide polymorphisms |
