Towards Quantifying Genetic Interactions Among Tumor Suppressor Genes in Breast Cancer

The effect of a gene's activity can be affected by the existence of another gene, termed as genetic interaction, or intragenic epistasis. Understanding genetic interactions among tumor suppressor genes will undoubtedly facilitate our understanding of the tumorigenesis and metastasis. With the wide-application of the next-generation sequencing of patient tumor samples, we have gained unprecedented resolution on the occurrence of mutations in each patient. Many previous studies have used odds ratio and Fisher's exact test to quantify such genetic interactions among tumor suppressor genes. Here, using the TCGA breast cancer dataset as an example, we have showed a potential problem for previous approaches and propose an approach that corrects for these bias. This study serves as a reminder that the interpretation of such large scale studies of mutational frequency data should proceed with reasonable caution.