A computational panel of pathological RAS mutants with implications for personalized medicine and genetic medicine

Abstract

The RAS proteins (KRAS, NRAS, and HRAS) play important roles in multiple diseases. This includes many types of cancer and the developmental syndromes collectively referred to as the RASopathies. There are many different RAS mutants that are found to drive these diseases. Mutant-to-mutant differences pose a challenge for personalized medicine. To investigate this problem, we extend our previously developed model of oncogenic RAS mutants to a total of 16 oncogenic mutants. We also extend our model to RASopathy associated mutants using data for 14 such RAS mutants. The model finds that the known biochemical defects of these mutants are typically sufficient to explain their elevated levels of RAS signaling. In general, our analysis finds that the oncogenic mutants are stronger than the RASopathy mutants. However, the model suggests that RAS signal intensities are spanned by the pathological variants; there does not appear to be a perfect separation between cancer promoting and developmental syndrome promoting mutants. Analysis of the panel also finds that the relative strengths of pathological RAS mutants is not absolute, but rather can vary depending on context. We discuss implications of this finding for personalized cancer medicine and for medical genetics. As genomics permeates clinical medicine, computational models that can resolve mutant specific differences, like the one presented here, may be useful for augmenting clinical thinking with their ability to logically translate biochemical knowledge into system level outputs of perceived clinical relevance.