output features calculation from various SNP predictor programs

Question

[arronslacey]

Hi - I want to collect information on SNPs that are often used to predict how damaging they are.

For example - using polyphen I can output not only the score, but all the features (MSA, domain regions etc) that were used to calculate the score by using the SNPFUNC parameter:

#!/bin/sh
curl \
  -F _ggi_project=PPHWeb2 \
  -F _ggi_origin=query \
  -F _ggi_target_pipeline=1 \
  -F MODELNAME=HumDiv \
  -F UCSCDB=hg19 \
  -F SNPFUNC= \
  -F NOTIFYME=[e-mail] \
  -F _ggi_batch_file=@snp.txt \
  -D - http://genetics.bwh.harvard.edu/cgi-bin/ggi/ggi2.cgi

which I find very useful as I can use these features to feed into a machine learning process. I am currently trying to do this with SIFT but cannot determine if this is possible. I have used VEP which can produce some SNP annotation, but VEP just querey's against a database that contains such information, rather than calculate it's own score.

Is anyone aware of other prediction tools that output their features as well as score?

thanks.

arron