Predicting synthetic lethal interactions using conserved patterns in protein interaction networks

In response to a need for improved treatments, a number of promising novel targeted cancer therapies are being developed that exploit human synthetic lethal interactions. This is facilitating personalised medicine strategies in cancers where specific tumour suppressors have become inactivated. Mainly due to the constraints of the experimental procedures, relatively few human synthetic lethal interactions have been identified. Here we describe SLant (Synthetic Lethal analysis via Network topology), a computational systems approach to predicting human synthetic lethal interactions that works by identifying and exploiting conserved patterns in protein interaction network topology both within and across species. SLant out-performs previous attempts to classify human SSL interactions and experimental validation of the models predictions suggests it may provide useful guidance for future SSL screenings and ultimately aid targeted cancer therapy development.Author summary: Our new algorithm SLant, uses artificial intelligence to help target future cancer drug research. In healthy cells tens of thousands of proteins work together forming large interaction networks. However, in cancerous cells genetic damage means that many of these proteins are disabled. Basic functions like DNA repair and signaling no longer work properly, and the cell replicates without proper control. Recent experience with breast cancer shows that gentler, more personalised therapies can be achieved by finding pairs of proteins which are ‘synthetically lethal’. The term means that the cell can cope if either one of the proteins does not work, but will die if neither of the proteins is functioning. Many synthetic lethal interactions are known, but there are many millions of potential pairs and finding new ones experimentally is difficult and time-consuming. SLant uses most of the experimental data that we have to identify the patterns in the protein interaction network associated with being part of a synthetic lethal interaction. By searching the network for proteins pairs that match these patterns, it can effectively predict new synthetic lethal pairs. The predictions were then checked against the rest of the experimental data. Our predictions are publicly available through the Slorth database.