Identifying Tier 1 Variants

[sgusev]

From the published work I've seen, there's a common theme of looking at Tier 1 variants in sequence data, specifically: non-synonymous/non-sense mutations, splice site variants, and frameshift indels. In some cases, this is easier said then done, requiring in-phase alignment and gene annotation. Does anyone know of a tool that or collection of tools that can pull out these features simply by knowing the physical position of a variant? I know trait-o-matic is heading in this direction but I don't believe they've tackled splice sites yet.

Thanks

[NGSfan]

*bump*

I am also very interested in this.

SeqWare looks like it could do this sort of automated variant annotation. It looks like a powerful suite of tools, but so far the documentation is very sparse and from my 5 minute look at the tutorial not so easy to just start up.

Anyone else know of something that can take in GTF annotation , say, and a list of variants and output frameshift/splice site/nonsynonymous annotation?

[sperry]

Originally Posted by NGSfan

*bump*

I am also very interested in this.

SeqWare looks like it could do this sort of automated variant annotation. It looks like a powerful suite of tools, but so far the documentation is very sparse and from my 5 minute look at the tutorial not so easy to just start up.

Anyone else know of something that can take in GTF annotation , say, and a list of variants and output frameshift/splice site/nonsynonymous annotation?

SIFT provides both SNP/indel annotation (http://sift.jcvi.org).

SeattleSeq (http://gvs.gs.washington.edu/SeattleSeqAnnotation/) allows for SNP annotation only.

Ensembl offers a 'SNP effect predictor' tool available via the 'Manage your data' link on the left-hand menu. (http://uswest.ensembl.org/Homo_sapiens/Info/Index)

You might also consider building your own annotation tool based on the Ensembl Variation API (http://uswest.ensembl.org/info/docs/..._tutorial.html)

[NGSfan]

thanks sperry for taking the time to list some SNP annotators and pointing me to some very useful links! I did not know any of these before!

This will definitely help me get started.

[NGSfan]

I just noticed that SeattleSNPs only handles SNP but not indels - is there any other software that can handle indel annotation?

Or even more simply - just take in SNP/indel coordinates (VCF or SAMtools pileup) plus gene annotations (GFF) and output HGVS annotation?

I'm amazed at the lack of tools for this...

[bioinfosm]

Thanks sperry.

I notice that seattlesnps is a good source, but anyone knows if it is downloadable, and if so, how much effort goes into setting up all the associated databases?

NGSfan, does SIFT work for you?

[Dongliang Ge]

Folks,

You may also want to check this out:

http://www.svaproject.org/

We develop SequenceVariantAnalyzer, or SVA, trying to annotate and visualize SNP, INDEL, and larger SVs.

[NGSfan]

Originally Posted by bioinfosm

Thanks sperry.

I notice that seattlesnps is a good source, but anyone knows if it is downloadable, and if so, how much effort goes into setting up all the associated databases?

NGSfan, does SIFT work for you?

SIFT is useful in that it can handle indels, whereas Seattle SNPs does only SNPs.

SIFT is also nice since you can download it and run it locally (I am about to try it).

However, SIFT does not automatically check if your variations are already in the dbSNP/1000genomes like SeattleSNP does.

[NGSfan]

Originally Posted by Dongliang Ge

Folks,

You may also want to check this out:

http://www.svaproject.org/

We develop SequenceVariantAnalyzer, or SVA, trying to annotate and visualize SNP, INDEL, and larger SVs.

Dongliang - I really liked your program and I think it is really answering the call to address this issue of variant annotation.

However, as a user I would like to have a command line only execution for a streamlined pipeline - do you think it could be done? I'd rather not have to click through everything, and use the GUI only for inspection of the results. You offer exporting of the tables - this is great, but command line only execution with tab delimited tables as output would be great. That way it could be easy integrated into a pipeline.

Oh yes and another thing to add to output HGVS format of variant annotation.

Ahh users, they are never happy are they? ;-)

Thanks for your great software.

Best,
German

[Dongliang Ge]

Hey German,

Thanks for trying out SVA and the comments! Also thanks a lot for the nice words.

Yes the suggest on command line tool is exactly my main focus now. I expect to release a command-line SVA tool sometime this week, along with some other updates and fixes.

I will include HGVS format too, but that will be a bit later. For the current release I only include a unique ID based on chromosome, location and type, I did not include HGVS which is transcript (or coding sequencing)-dependent. You are right it should be part of the annotation output for sure.

I hope this tool can be of any help in your research. And please let me know if you have other comments.

Regards,

Dongliang

Originally Posted by NGSfan

Dongliang - I really liked your program and I think it is really answering the call to address this issue of variant annotation.

However, as a user I would like to have a command line only execution for a streamlined pipeline - do you think it could be done? I'd rather not have to click through everything, and use the GUI only for inspection of the results. You offer exporting of the tables - this is great, but command line only execution with tab delimited tables as output would be great. That way it could be easy integrated into a pipeline.

Oh yes and another thing to add to output HGVS format of variant annotation.

Ahh users, they are never happy are they? ;-)

Thanks for your great software.

Best,
German

[lshen]

I used Ensembl API to annotate my SNPs from mouse.

http://pre.ensembl.org/info/docs/api...ion/index.html

So far, only using it to:

Calculated data:

* Consequence of variations in transcripts (synonymous, non-synonymous,...)

[Dongliang Ge]

Dear German,

A command line tool is just released in a newer version SVA 1.01beta, according to yours and others' suggestions.

Here is a brief introduction to this tool:

http://www.svaproject.org/usecmd.php

Here is how to update your current SVA package (if you are running a standard release instead of a Lite Evaluation release - a Lite Evaluation release cannot run an annotation job):

http://www.svaproject.org/versionupdate.php

Again I hope this tool can be of any assistance.

Best regards,

Dongliang

Originally Posted by NGSfan

Dongliang - I really liked your program and I think it is really answering the call to address this issue of variant annotation.

However, as a user I would like to have a command line only execution for a streamlined pipeline - do you think it could be done? I'd rather not have to click through everything, and use the GUI only for inspection of the results. You offer exporting of the tables - this is great, but command line only execution with tab delimited tables as output would be great. That way it could be easy integrated into a pipeline.

Oh yes and another thing to add to output HGVS format of variant annotation.

Ahh users, they are never happy are they? ;-)

Thanks for your great software.

Best,
German

[NGSfan]

Dongliang,

Wow! awesome! that is great news. We have a new data set coming in and so this will be a great opportunity to try it out.

Thanks a lot for adding this! I'm looking forward to using it!