Learning Altered Disease Biology in Patients with Network Models and Clinical Data Analysis

Already, we understand that drugs can have multiple effects on disease symptoms and side-effects, and further that they often act through on- and off- target protein binding, yet we underutilize this information at initial drug design or when prescribing. We previously discovered that reclassifying drugs by their effects on proteins downstream of known drug targets can better predict clinical effects. In this study, we developed a novel ALS-focused protein interaction analysis using datasets collected from patient samples, the public domain, and consortia. We used these pathways to design new drug classes, and assessed class-level effects on ALS overall survival in a retrospective analysis of real-world data. With complementary in vitro analysis, we observed that drugs that induce complement activation are associated with increased risk for death, implicating this pathway as a viable therapeutic target for treating ALS. We further extended our approach to search for six protein classes, three of which, including multiple chemokine receptors, had more significant effects on ALS survival, suggesting that targeting proteins such as chemokine receptors could be advantageous for these patients. We also tested four drug classes in Parkinson’s and Myasthenia Gravis. We again recovered effects for drugs associated with complement activation and discovered that drugs associated with chemokine receptors had stronger effects on overall survival. Taken together, these results demonstrate the utility of a network medicine approach for testing novel therapeutic effects using real world data and suggest actionable therapeutic strategies for multiple neuroinflammatory diseases. In addition to insights about neurodegeneration, this suggests a framework for understanding altered disease biology in patients taking medications consistently.