In my project I am analysing SNPs in cancer datasets. I want to compare the allele frequencies occurring in "my" datasets with those in other populations. For this, I am using 1000 genomes, which works fine in a lot of cases.
However, in some cases, I will find a SNP at a position, where there is no frequency data available from 1000 genomes. Quite often, these SNPs do though appear in dbSNP.
Is there some way how I can estimate the frequency of such alleles in other populations? Maybe by using a "panel of normals", as suggested in GATK? But which dataset would be used for such kind of a panel?
Thank you for any suggestions and help :-)