Question

Variant Effect Prediction In Non-Human Organisms

2

Entering edit mode

12.1 years ago

Rubal7 ▴ 40

Hello everyone,

I'm working with Rattus norvegicus and want to know if people can recommend good approaches to predict the impact of non-reference variants. I have used ensembl's variant effect predictor and snpeff. I was wondering if people had opinions on which method is best. Also I'd love to know which of my variants are effecting typically highly conserved bases(in mammals). I believe some software does this for Human variants but is not applicable to non-human species. If anyone knows of a systematic way of doing this other than checking manually on a genome browser that would also be really helpful.

Thanks in advance,

Rubal

variant variant genome genome • 2.8k views

ADD COMMENT • link updated 12.1 years ago by Zev.Kronenberg 12k • written 12.1 years ago by Rubal7 ▴ 40

score 2 · Answer 1 · 2012-03-08

2

Entering edit mode

12.1 years ago

Zev.Kronenberg 12k

VAT in the VAAST pipeline can annotate effects of SNVs/INDELs in non-model organisms. All you need is a GFF3 file and a VCF. Let me know if I can help you.

ADD COMMENT • link 12.1 years ago by Zev.Kronenberg 12k

1

Entering edit mode

thanks, I'll look into this today

ADD REPLY • link 12.1 years ago by Rubal7 ▴ 40

0

Entering edit mode

I've not worked with GFF3 before. There should be one standard GFF3 file for my species (Rattus norvegicus)right, rather than that I need to create a new one based on my data? Thanks

ADD REPLY • link 12.1 years ago by Rubal ▴ 350

0

Entering edit mode

Also it seems that some aspects of VAT use information from primates to predict important changes in non-coding variants. Have people actually used this for non-human organisms?

ADD REPLY • link 12.1 years ago by Rubal ▴ 350

0

Entering edit mode

I have used it for mouse and pigeons. Works well. Found at least one gene controlling a phenotype.

ADD REPLY • link 12.1 years ago by Zev.Kronenberg 12k

0

Entering edit mode

Hi Zev, so I am working on the VAAST pipeline now. I am converting my VCF files to GVF using the vaast_converter script. As I am using non-human organism, I can't specify hg18 or hg19 for the build option. I tried specifying the rat GFF3 file but I get the error:

unknown_build_option_value : sequence-region pragmas will not be set: $build

Do you know if I am specifying the wrong thing or if there is a way I modify my GVF files later so they have the correct annotations?

Many thanks,

Rubal7

ADD REPLY • link 12.1 years ago by Rubal ▴ 350

0

Entering edit mode

Rubal, Sorry I am late in getting back to you. Since you are working on non human you need to specify a build file. I will send you one tomorrow morning. Please join our VAAST user mailing list. You will get answers much faster that way. We are excited that people are using VAAST for non human data and will answer all your questions. From GVF you are only two steps from running VAAST.