Abstract
Predisposition to non-familial, sporadic cancer is strongly influenced by multiple tumor susceptibility genes (TSGs), each with apparently minor effects on the cancer phenotype. Sequence analysis of the human genome has yielded numerous single nucleotide polymorphisms (SNPs), raising the expectation that new low-penetrance TSGs will be identified that can be used to estimate an individuals cancer risk. However, mouse models for human cancer showed that the effects of many low-penetrance TSGs are highly variable due to their involvement in epistatic interactions. Together, these interacting TSGs form large molecular networks, which represent cancer-associated biological modules that influence the tumorigenic process. As a consequence, although allelic variation in one TSG on a permissive genetic Background can have major effects on tumor development, the net effect of allelic variation in multiple interacting TSGs remains hard to predict. Therefore, the predictive value of SNP-analysis to estimate an individuals cancer risk will be restricted to those TSGs that exhibit single-gene effects. New strategies need to be developed to evaluate cancer risk associated with biological modules that are influenced by TSG-networks.