Functional annotation of SNPs and genes containing nonsynonymous SNPs

Question

[reza]

hi everyone

first question: i have a vcf file resulted from samtools, belonging to a mammalian that assembled in scoffold level. I annotated it and now want to extract non-synonymous SNPs and found genes containing ns SNPs. how i can extract sequences of genes that have ns SNPs to blast them and finding gene names.

second question is: how can i extract indels that located in genic region (in annotated vcf file) and get its length to plot it?

thanks in advance

[cpad0112]

try VEP and tutorial is here: http://www.ensembl.org/info/docs/tools/vep/script/vep_tutorial.html. Once annotated, you can filter the variants (Nonsynonymous) with full annotation. For filtering nonsynonymous variants, follow the tutorial here: http://www.ensembl.org/info/docs/tools/vep/script/vep_filter.html. Note that VEP can be customized to annotate variants only in coding regions and look at the VEP options for this.

[reza]

thanks for your answer but my under study animal is not in ensembl. i used snpeff to annotation.

[cpad0112]

Try snpsift on snpeff output: http://snpeff.sourceforge.net/SnpSift.html

Example code (modified from manual) to filter missense variant :

java -jar SnpSift.jar filter "ANN[*].EFFECT has 'missense_variant'" snpeff_annotated.vcf  >  fitlered_output.vcf

For indel filtering:

java -jar SnpSift.jar filter "(( exists INDEL )" snpeff_annotated.vcf > filtered.vcf

[reza]

thanks, it is helpful, i try it with "( EFF[*].EFFECT = 'NON_SYNONYMOUS_CODING' )" and it worked.

[reza]

your suggested way worked for extraction one effect but when i try it for several effect (below command), it did not worked.

java -jar SnpSift.jar filter "( EFF[].EFFECT = 'NON_SYNONYMOUS_CODING' )" & "( EFF[].EFFECT = 'STOP_GAINED' )" & "( EFF[*].EFFECT = 'STOP_LOST )" snpeff_annotated.vcf > fitlered_output.vcf

how can i extract several effects simultaneously?

[h.mon]

For your second question, bedtools intersect can extract SNPs / indels intersecting you annotation, if you have it on bed or gff format. You can easily get gene length from the bed / gff as well.