Finds single-base differences (substitutions) between sequence data from tumor and matched normal samples.