how to predict nucleotide mutation effects on a genome-wide scale?

[jnfass]

I'm familiar with (but have not used) ParseSNP (http://www.proweb.org/parsesnp/), a tool that calculates the effect (frame shift, truncation, etc.) of nucleotide mutations (indel, substitution) on genes, given the gene model. However, I'd like to do this on a genome-wide scale: load in a long list of mutations, and return protein effects if the mutation is within a gene. And I don't really think the people hosting ParseSNP would appreciate me bombing their servers by submitting entire chromosomes (of Arabidopsis thaliana)! Maybe it wouldn't be a problem ... I'll ask them.

But, in the meantime, are there more convenient, fast tools, for which the source is available, that do this? Any suggestions or advice would be appreciated.

Thanks,
~Joe

[krobison]

There are several packages in the Wiki which are in this space

[swbarnes2]

I just downloaded one called annovar. I'm working in mouse, but A.thaliain ought to be available too. I followed the manual, downloaded the mouse databases, converted my vcf file to it's format using the included script, and ran it. I got output that looks like this:

intronic Col15a1 4 47234472 47234472 A G 99
intronic Col15a1 4 47244368 47244368 A G 99
intronic Col15a1 4 47299820 47299820 - AGAAGAAGAAGAAG 99
intergenic Tgfbr1(dist=6268),Alg2(dist=48641) 4 47434064 47434064 T C 99
intergenic Sec61b(dist=88256),Nr4a3(dist=479757) 4 47584361 47584363 GCT - 99

The next step is to automatically check intronic mutations to see if they would cause cryptic splice sites, but I don't know of a site that will do that in batches.

[jnfass]

Thanks!

Thanks krobison, swbarnes,

I checked out your search results link, krobison ...
GAMES and MU2A seem to be limited to human (similarly for SNPnexus, which I just stumbled across), and MU2A is a big install job involving TomCat ...
VIP is for 454 reads (I've got Illumina).
VariantClassifier from JCVI looks promising, but I think I'll try Annovar (thanks swbarnes!)

Appreciate the tips,
~Joe

[krobison]

GAMES doesn't appear to be human-specific, but is oriented towards genomes in the UCSC Browser, which unfortunately doesn't include Arabidopsis.

[laura]

The ensembl system comes with a variant effect predictor which you might find useful

http://plants.ensembl.org/tools.html

[jnfass]

Thanks Laura! - Ensembl's Variation Effect Predictor worked perfectly, and the online version is set up for Aribidopsis and about 8 other plants, so it didn't even require any hacking.

We talked to Kai, who wrote (or is maintaining Annovar), and he added a FAQ question about using a genome that's not in UCSC's database (here), but we didn't fully explore how easy that would be, having found Ensembl's tool first.
Thanks again, all ...
~Joe