Medicine is targeted to standards of care, providing a “statistical probability of health”; however, considering individual variability, a “treatment good enough for everybody is not perfect for anyone”. The goal of personalized medicine is to provide an objective basis to quantify individual variability, emphasizing the customization of healthcare, with all decisions and practices being tailored to individual patients.
Our goal is to develop algorithms and software tools that, starting from experimental data, automatically generates a protein-protein interaction network integrated with protein phosphorylation levels, thus leading to experimentally-compliant network topological properties. The procedure will be critical to mark out individual subjects using network-based bio-markers, obtained merging the information derived from high-throughput experiments with network topological analysis.
We previously laid the foundation of the method (NetBiomarkers 2015), identifying protein phosphorylation levels as the best experimental data to be integrated with network topology and defining the pipeline to perform the automated analysis on the enriched network.
A multithread analysis procedure and software allowing comparison of multiple networks will be implemented, letting parameters computation of several subjects. A preliminary design of a dedicated chipset running the new algorithms will also be performed, thus strongly increasing the efficiency of the entire procedure with a tremendous performance gaining and allowing analysis of hundreds of networks in few minutes.
The new software will be applied to hundreds of phospho-proteomics data sets provided by Kinexus. By applying network analysis methods to such data sets we will improve the knowledge of biological processes, confirming target proteins and suggesting new ones, discriminating false and positive negatives, thus allowing identifying parameters and values better characterizing each condition to detect specific changes between networks obtained in different biological contexts (for instance, healthy vs non-healthy subjects).