Exome IV: Variant annotation and filtering

Course: HT17 Analysis of next generation sequencing data (SC00024)

---

Now we need to annotate our variants in terms of genomic context, aminoacid change and effect, previous mutation knowledge, etc. so we can evaluate and filter mutations that are irrelevant.

Variant annotation (Annovar)

ANNOVAR is a tool that annotates genetic variants. The overall procedure is to create an input file from the VCF file. Use $convert2annovar using -format vcf4 as a required flag.

Then we annotate the positions with the use of several databases. This is a tricky command line so use:

$table_annovar.pl <your_variant_file> $humandb \
   -buildver hg19 \
   -protocol dbsnp_137_chr11_gatk,refGene,ljb2_all,1000g2012apr_all,cosmic65 \
   -operation f,g,f,f,f \
   -outfile <your_output_file> \
   --nastring - \
   --remove

Q. What kind of information are we adding to our mutations? Have a look at the file with .hg19_multianno.txt extension. Here you can find some information.

Now paste the file you created from the VariantsToTable tool and the AS_2*hg19_multianno.txt. Save it in a file with extension .tsv. This file now has statistics and annotations for the mutations, if you have a look here is a description of some columns:

Column	Description
AC	Allele count
GT	Genotype
AD	Allele depth
DP	Depth
SIFT	D = damaging, T = tolerated
HDIV	D = probably damaging, P = possibly damaging, B = benign
HVAR	D = probably damaging, P = possibly damaging, B = benign
LRT	D = deleterious, N = Neutral , U = Unknown
MutationTaster	A=disease_causing_automatic, D=disease_causing, N=polymorphism, P=polymorphism_automatic
phyloP100way	the higher the value the more conserved is the aminoacid in other organisms

This is how you read the AAChange column:

PGM1:NM_001172818:exon2:c.A316G:p.I106V,PGM1:NM_002633:exon2:c.A262G:p.I88V

PGM1 -> Gene name
	NM_001172818 => Transcript ID_1
		exon2 => exon where the mutation is found
			c.A316G => in the DNA, at position 316 the reference base A is changed to an G
			p.I106V => in the protein, at positon 106 the reference aminoacid I is changed to a V
PGM1 => Gene name
	NM_002633 => Transcript ID_2
		exon2 => exon where the mutation is found	
			A262G => in the DNA, at position 262 the reference base A is changed to a G
			p.I88V => in the protein, at position 88 the reference aminoacid I is changed to a V

Variant filtering

With all this information, it is possible to narrow down the list of mutations to a handful of SNPs. Think of a filtering strategy and filter your SNPs.

Q. Which are your candidate genes? What was your filtering strategy?

Once you have your list, visual inspection of your candidates helps in assessing them further. Have a look at your candidates in IGV, you may want to load both of your VCF files as well as the BAM files

Q. How are the mutations distributed? Are they all in exons? How do you differentiate between heterozygous and homozygous mutations? Some mutations are slightly faded, what does this mean?

Q. Could you filter out some of your candidates with your visual inspection? Which are you final candidate genes?

---

Previous: Exome III: Alignment processing and variant discovery

---

Created by Marcela Dávila