MEDPNC is developing a diagnostic test using a quantitative approach that maps each tumor accurately and unambiguously based on its molecular disruptions, in a way that allows adapting drug combinations to treat each patient optimally. Existing solutions in the market are based on the analysis of overexpression of oncogenic biomarkers only and lack important information regarding the interplay between these markers. Without the accurate mapping of the protein-biomarkers within an altered subnetwork, their relationships with other proteins as well as other subnetworks in the unbalanced state may be overlooked, leading to wrong decisions regarding therapy combinations. Consequently, the selected therapeutic regimen may lack long-term efficacy resulting from partial targeting of the tumor imbalance. Large-scale cancer datasets have been rapidly accumulating over the past years. The challenge has now become to find a way to obtain useful information from this data that will allow for the determination of which combination of approved drugs is best suited to treat each specific tumor. Various statistical analyses are being developed to extract significant signals from cancer datasets. However, tumors are still being assigned to predefined categories, conceptually contradicting the vast heterogeneity that is known to exist among tumors, and likely overlooking unique tumors that must be addressed and treated individually. MEDPNC uses a thermodynamic–based information theoretic approach that, rather than searching for what makes a tumor similar to other tumors, addresses tumors individually. The algorithm maps every malignancy precisely and unbiasedly according to the protein network reorganization in each tumor, namely the patient-specific signaling signature (PaSSS). This PaSSS contains unbalanced molecular processes in each tumor which should be targeted, in order to return the processes to the stable state. The obtained PaSSS, contains essential information for predicting the response of the tumor-specific network to different therapeutic modalities, and allows the rational design of smart, personalized drug combinations. Thus, this approach is superior to more common statistical analyses used currently which extrapolates knowledge about a subpopulation of patients to the entire population. MEDPNC’s product uses proteomic datasets obtained from multiple patients suffering from several types of cancer and recommends the optimal treatment combination based on each patient’s PaSSS and not on the anatomical origin of the tumor. Tumors from different origins may share the same PaSSS. Thus, the treatment is designed to target the particular PaSSS instead of using the conventional treatment focusing on particular cancer types and the anatomical site of the tumor. For the validation of this approach MEDPNC researched several cancer types such as breast and melanoma, and showed that the therapies predicted achieved significantly higher rates of killing than the clinically prescribed treatments. Moreover, MEDPNC demonstrated that within the same type of cancer, e.g melanoma, different treatment combinations should be used according to each PaSSS in order to achieve selectivity and specificity of the treatment. Thus, this research scheme has a high potential to be translated rapidly into a computational approach that is expected to fill the currently unmet need for an accurate and predictive framework for the rational design of optimal anti-cancer personalized therapy.