Abstract: The disclosure features methods of identifying specific drug candidates for disease treatment, the methods including: for pairs of associated expressed proteins in a protein-protein interaction network for a plurality of biological samples, comparing relative concentration values of the associated proteins in each of the biological samples to identify outliers among a distribution of the relative concentration values, identifying a set of proteins involved in dysregulated protein-protein interactions in the network based on the outliers, and predicting an efficacy of one or more drug candidates from among a set of drug candidates for treating a disease based on the identified set of proteins and information about one or more of proteins, protein complexes, biological pathways, and functional modules targeted by the set of drug candidates.

Abstract: The disclosure features methods of identifying protein-protein deregulation that include: generating a basal protein-protein interaction network for a plurality of biological samples, the network featuring a set of proteins expressed in the biological samples and concentrations of each member of the set of expressed proteins in each of the biological samples; identifying two associated expressed proteins in the network; for the two associated expressed proteins, comparing correlated relative concentration values of the two proteins in each of the biological samples to identify outliers among a distribution of the relative concentration values; and identifying members of the plurality of biological samples in which deregulation of the two associated expressed proteins occurs based on the outliers.

Abstract: The present invention provides methods and compositions for increasing tumor antigenicity by increasing the level of expression of phosphoneoantigens in cells (e.g., restoring the expression of a phosphoneoantigen on the surface of a cancer cell). The invention also provides methods and compositions for enhancing recognition of phosphoneoantigens by immune cells.

Abstract: The disclosure features methods of identifying specific drug candidates for disease treatment, the methods including: for pairs of associated expressed proteins in a protein-protein interaction network for a plurality of biological samples, comparing relative concentration values of the associated proteins in each of the biological samples to identify outliers among a distribution of the relative concentration values, identifying a set of proteins involved in dysregulated protein-protein interactions in the network based on the outliers, and predicting an efficacy of one or more drug candidates from among a set of drug candidates for treating a disease based on the identified set of proteins and information about one or more of proteins, protein complexes, biological pathways, and functional modules targeted by the set of drug candidates.