Questions About Annotation Of Mouse Exome Variants Data By Annovar

Question

[Tonyzeng]

HI, my mouse variant data (VCF format) has been processed by recalibrating, snp calling and filtering by GATK package.

Now it is time to annotate this variant data by ANNOVAR package. I processed it as follows,

1) convert the VCF file format to the annovar file format:

$convert2annovar.pl --format vcf4 --includeinfo snps.filtered.vcf > snps.annovar &

2) I have one package - mouse mm10 annotation package that is downloaded from Illumina resource ( based on UCSC March 6, 2013)

This package contains 3 different files including "Genes file" which is consisted of ChromInfo.txt, genes.gtf,refFlat.txt.gz,refGene.txt and refSquSummary.txt, "Small RNA file" includes mature.fa and precursor.fa and "variation file " has snp137.txt.gz.

From an example of using annovar in human the command is like this:

[kai@beta ~/]$ annotate_variation.pl ex1.human -buildver hg19 humandb/ (copied and modified from http://www.openbioinformatics.org/annovar/annovar_gene.html)

For the mouse variants annotation, I was totally lost now, my questions are

1) My annotation in my computer seems a little bit complex, which one I use? or use just the whole package? To make my analysis straight forward, Where to download the mm10 annotation data WHICH works EXACTLY for annotation of ANNOVAR?

2) What is the correct or PERTINENT commands for my mouse exome annotation by ANNOVAR? ( the example above for human is obviously not for my mouse data)

Thank you!

Answer 1

[Ashutosh Pandey]

1) If ANNOVAR doesn't provide the mm10 data then you need to manually create it. I am sure it provides you the commands to how to build the annotation database from scratch. Most of the information for mm10 can be retrieved from UCSC ftp site. First of all you need to know what information you want. If you just want to annotate SNPs and Indels to find out which genes and which exons have been affected that would be easy to do. But if you want to see if any of your variants are located in a particular TFBS, then you need to supply a bed file for that particular TF while you are creating the annotation database.

2) Everything is mentioned in the manual. Take some time to go through it. Asking trivial questions on forum just to finish some task won't help you in long run. Please don't take it otherwise, I only want to help you.

[Tonyzeng]

thank you! ashutoshmits

I've downloaded mmdb according to Annovar manual. However, when I run

$ summarize_annovar.pl --buildver mm10 snps.annovar mm10db -outfile snps &

It showed

$ Error in argument: the --buildver argument can be 'hg18' and 'hg19' only Usage:

[Ashutosh Pandey]

Did you follow steps mentioned in this webpage (http://www.openbioinformatics.org/annovar/annovar_db.html) carefully to create a new mm10 database ?

[Tonyzeng]

Sorry I did not keep updates that summarize_annovar.pl was obseleted already.

Answer 2

[Charles Warden]

I believe summarize_annovar.pl is a wrapper that is specifically used for common human annotations and won't work with other genomes.

When I want to provide GWAS catalog annotations (which is a human annotation track not included in summarize_annovar.pl), I use annotate_variation.pl. For the GWAS catalog annotations in humans, the command looks like this (for you, buildver = mm9 or mm10, --dbtype = custom mouse database, and the command ends with the annotation folder - so, "humandb" in my example is probably like "mm10db" in your code - where you downloaded / created your mouse annotations):

/annovar/path/annotate_variation.pl -regionanno -out output_file.ext -buildver hg19 --dbtype gwascatalog input_file.ext /path/to/humandb

I think you'll need to do something similar with any mm10 tracks that you create / download.