snp analysis using a software tool

Question

[anusha.sunkum]

I have variant data obtained from a tool that give variants in with reference to a genome. I had tried to analyse them, like their position effect on amino acid,...etc using snpeff, but this tool needs it in a variant call format which is obtained from sequencing experiments only. Can u suggest me any tool to analyse the snp data...or how can snpeff can be used in this case

Thank you

[Devon Ryan]

What's the format that you have now? It's difficult to determine what you actually have given what you've written.

[anusha.sunkum]

I have gff3 format with me.

Thanks for your reply

[anusha.sunkum]

Okay let me be clear. I already have snp data with insertions, deletions, indels in gff3 format.To feed it into snpeff and analyse it I need it in varient call format which needs quality, filter, information in input.so in gff3 format I am not able to make out what is quality and filter.

[Devon Ryan]

There's no standard way to represent variants in gff3, since that's not it's purpose. You'll just have to write a short script to convert it to VCF.

Answer 1

[Bert Overduin]

The Ensembl Variant Effect Predictor (VEP) takes many different formats as input and is available through a web interface, as a downloadable script and also through the Ensembl REST API.

[Bert Overduin]

The default input format for the VEP is a simple is whitespace-separated format:

• chromosome - just the name or number, with no 'chr' prefix
• start
• end
• allele - pair of alleles separated by a '/', with the reference allele first strand - defined as + (forward) or - (reverse). identifier (optional) - this identifier will be used in the VEP's output. If not provided, the VEP will construct an identifier from the given coordinates and alleles.

[anusha.sunkum]

Does it work for plants?

[Bert Overduin]

For the ones that are in Ensembl Plants, yes. For the VEP, just have a look at the Ensembl Plants Tools page.

Answer 2

[Pablo]

Just transform your format to VCF and feed it into SnpEff.

[anusha.sunkum]

Is there any tool to transform or should I do it manually?

Thank you

[Pablo]

Honestly, if I were you, I would just re-run the variant calling phase using a caller that outputs VCF format (which is almost every variant caller I know, since that is the standard).

Keep in mind that GFF3 is certainly not the right format for varaints, so there must be something very "special" (possibly wrong?) in your pipeline. If you transform GFF3 to VCF, you can leave some fields empty (such as FILTER, QUALITY and even INFO) just use '.' to represent 'no value'.

So the fields in VCF are tab separated:

CHROM
POS
ID       (can be empty, '.')
REF
ALT
QUAL (can be empty, '.')
FILTER (can be empty, '.')
INFO (can be empty, '.')

As you can see, it should be quite easy to get your data into VCF format.