In cancer, tumors accumulate genetic mutations at a strongly enhanced rate. Some of these
mutations are shown to drive cancer development. In most of the cases, these so-called cancer
driver mutations correspond to mutations in proteins. Proteins interact with each other to fulfill
their function. Often, cancer driver mutations are found exactly at the contact interface between
two interacting proteins. This can lead to alterations of this protein-protein interaction, which can
contribute to cancer development. We here develop a method to map the complete interaction
interfaces between proteins at unprecedented throughput. We apply this method to the proteins
p53, PTEN and PIK3R1, three major targets of cancer driver mutations. This will show how
different mutations in these proteins affect the interactions with a large number of interaction
partners and will help to explain how different mutations in the same protein can be linked to
different cancer types, different disease progression or different response to therapy. The results
of this work can support the classification of cancer driver mutations and can lead to better
prognosis and tailored therapy in cancer treatment.