We considered all variants in a cancer sample that passed the statistical assessment described above as potential somatic mutations. We then retained the mutations for which the corresponding position in the germline sample was: 1) highly covered (> 10 reads); 2) not called mutant using the same statistical assessment; 3) did not show strong evidence of an alternative allele (> 20%) that failed statistical assessment based on filter 4. Therefore, while most germline mutations will fall under criterion 2, this last criterion identifies rare cases of obvious germline mutations (20% allele frequency), which failed the statistical assessment in the blood sample.
