From In Silico Identification Of Deleterious Snps To In Vitro Validation Of Same

Question

[Dataminer]

Hi!

I have seen and heard a lot of people using SIFT and PolyPhen and other computational tools, in order to identify potentially functional polymorphisms.

But, is their any example, where these tools have actually lead to an in vitro verification of the same SNP(which was identified as deleterious by these tools)?

I was just wondering.

Anyone, ever came across such an article.

Thank you

[Larry_Parnell]

Just a small note of clarification - tools just as SIFT and others (eg, transcription factor binding site predictors) may identify potentially functionally relevant polymorphisms. Potential is a kayword here in teh absence of molecular biology and biochemistry.

[Dataminer]

Thanks, Larry. I will make the correction.

Answer 1

[Dpsguy]

SIFT and Polyphen are mostly used for doing the opposite of what you ask: computationally verifying associations that are found in GWAS or epidemiological studies. However, you could go through the following articles to see if this is what you want:

An Evolutionary Perspective on Single-Nucleotide Polymorphism Screening in Molecular Cancer Epidemiology

In this study, we surveyed odds ratios detected for 46 SNPs in 39 different cancer-related genes from 166 molecular epidemiological studies. The conservation levels of amino acid that these SNPs affected were calculated as a tolerance index by comparing sequences from different species (SIFT). Our results provide evidence of a significant relationship between the detected odds ratios associated with cancer risk and the conservation levels of the SNP-affected amino acids (P = 0.002; R2 = 0.06).

Predicting deleterious amino acid substitutions

Comparison of SIFT with BLOSUM62 Predictions on LacI Mutation Data: Of substitutions predicted to be deleterious by SIFT, 66% will yield a deleterious phenotype experimentally by the β-galactosidase assay.

Comparison of SIFT with BLOSUM62 Predictions on HIV-1 Protease Mutation Data: Out of 215 substitutions predicted by SIFT to be deleterious, 85% give an affected phenotype using the protease assay by Loeb et al. (1989).

Pattern of Sequence Variation Across 213 Environmental Response Genes

Of the 23,443 SNPs found in the 213 candidate genes, 541 nonsynonymous cSNPs were identified. We evaluated these cSNPs to identify variants with potential functional consequence by using two homology-based tools: SIFT (Ng and Henikoff 2003) and Polyphen (Sunyaev et al. 2001; Ramensky et al. 2002). Fifty-seven SNPs were identified as potentially deleterious by both of these approaches. For a subset of these variants (n = 36), we were able to identify a functional domain associated with the polymorphism by using annotation from the Human Gene Mutation Database. Notably, we predict intolerant cSNPs in 31 genes with no entry in the Human Gene Mutation Database. Seven of these predicted intolerant cSNPS have allele frequencies >5% in the PDR, four of which are not listed in the Human Gene Mutation Database. Of the 57 cSNPs predicted to be intolerant, seven are reported to be associated with a known phenotype.

You may also go through this excellent review article: Assessing the function of Genetic variants in candidate Gene association studies

Answer 2

[Alex Paciorkowski]

I would say SIFT and PolyPhen are commonly used by any lab interested in gene discovery. These tools are used to evaluate variants found during either Sanger or next-gen sequencing projects. Labs often decide which variants to pursue with functional validation studies based upon scores provided by these algorithms.

SeattleSeq can also provide PolyPhen scores as part of its annotation of NGS variants.

As for papers that detail the use of these tools in the way I think you are describing, you can check out Endele, et al as just one example. A search of PubMed for any of the recent human genetic syndromes where pathogenic mutations have been found will come up with many others, I'm sure.