Method of providing a treatment to a subject based on a dynamic pathway map

A method of generating a dynamic pathway map (DPM), comprising:

• accessing a pathway element database storing at least one pathway including a plurality of pathway elements comprising genes, RNA, and proteins;
• associating a first pathway element of the pathway element database with at least one known attribute corresponding to a gene, an RNA, or a protein by connecting the first pathway element to the at least one known attribute using at least one edge in a factor graph;
• associating a second pathway element of the pathway element database with at least one assumed attribute corresponding to a gene, an RNA, or a protein by connecting the second pathway element to the at least one assumed attribute using at least one edge in the factor graph;
• cross-correlating the first pathway element, the second pathway element, and additional pathway elements for the at least one pathway using the edges in the factor graph;
• assigning, based on the cross-correlating, an influence level of the first and second pathway elements relative to activity or expression of each other or the additional pathway elements for the at least one pathway using the at least one known attribute and the at least one assumed attribute, wherein the influence level corresponds to an edge of the factor graph;
• converting each pathway of the at least one pathway, and the influence level of the first and second pathway elements for the at least one pathway, to a probabilistic pathway model having each interaction in each pathway represented as an edge in the factor graph of the probabilistic pathway model, with the influence level of the first and second pathway elements comprising one of a negative value or a positive value, the negative value and the positive value correlated to the corresponding edge of the factor graph associated with the at least one pathway;
• storing the probabilistic pathway model, additional pathway models, and a plurality of measured attributes for a plurality of elements based on a genome-scale assay of a biological sample, each instance model having edges between pathway elements comprising genes, RNA, and proteins, with the edges having influence levels;
• deriving the DPM from expectation maximization of the probabilistic pathway model and the plurality of measured attributes for the plurality of elements based on the genome-scale assay of the biological sample, as well as the additional pathway models, to determine influence levels across the pathway models, the DPM inferring probabilistic reference pathway activity information for pathway elements of a particular pathway;
• and
• identifying that an activity or an expression of a particular pathway element of the particular pathway is altered for the biological sample compared to a control based on pathway inferences in the DPM, wherein at least one method operation is executed by a processor, and wherein the biological sample was obtained from a subject;
• and
• providing a treatment to the subject based on the activity or the expression of the pathway element of the particular pathway being altered compared to the control, wherein the treatment is administration of a pharmaceutical.