Figure 1

Schematic of the MultiP pipeline (A). The input into the model is data from multiple platforms for the same cohort of patients. (B) For a particular sequence of platforms, the MultiP algorithm uses machine learning to classify patients into a positive, negative, or uncertain class. Patients in the uncertain class are passed onto the next platform and the process repeats until the final platform. (C) Candidate pathways corresponding to different orders of platforms can be compared and optimized on different criteria. (D) A suite of tools for visualizations and summaries of the constructed pathways are provided.