[ensembl-dev] Fwd: Variant effect predictor

[Mark Aquino]

> 
> 
> I'm attempting to validate the sift and pph2 scores I've gotten running the variation effect predictor on my data set but neither the SIFT nor PPH2 web applications return the same results.  Have the precomputed scores on ensembl become outdated due to better alignments or is something else at play here?  
> 
> Example output line from VEP.pl:
> 11409   1:1599812       T       ENSG00000189339 ENST00000378662 NON_SYNONYMOUS_CODING   1695    934     312     V/I     Gtc/Atc rs76114385      SIFT=tolerated(0.47);PolyPhen=possibly_damaging(0.272);Condel=neutral(0.312)
> 
> The SIFT result from the website using dbSNP returns a score of 0.68 and Polyphen returns a score of 0.511.
> 
> I'm interested in knowing the reason for the discrepancy.
> 
> Thanks,
> Mark Aquino
> 
> 
>