Is there any better method to find significant SNVs ?

Question

[mangfu100]

Hi.

I have worked with normal-tumor SNVs results by using VarScan2 and Mutect.

So I have total 500 candidate somatic nucleotide variants.

but problems is that there is so many snvs in my data. I need significant snvs which is related to cancer.

So I think that i have to search databases liek dbSNP or 1000genome.

so how to utilize above databases well? or Is there anything better method to find somatic variants in large sets?

Answer 1

[poisonAlien]

Go and paste your variants into oncotator.

The resulting maf has dbDSNP, COSMIC, TCGA, mutsig annotation data, which should be enough for you to begin with.

Also take a look at annovar if you want to 'search databases liek dbSNP or 1000genome'

[Charles Warden]

+1 for ANNOVAR. I would recommend checking the ClinVar annotations. I've found that COSMIC includes a lot of variants, including ones with very high population frequencies (say >70%). In other words, ClinVar has a much smaller list of cancer variants than COSMIC (and the ClinVar cancer variants seem more likely to be rare variants).