CLASSIFICATION OF PATIENTS WITH PRIMARY C3 GLOMERULOPATHIES/ IMMUNE COMPLEX-MEDIATED MEMBRANOPROLIFERATIVE GLOMERULONEPHRITIS
We implemented a computational tool that allows assignment of patients to a cluster using a set of genetic, laboratory and clinical parameters. The tool provides essential data portraying analysis results of the cohort of 295 MPGN patients including, among others, the patients’ variable landscapes and two-dimensional patient maps. It allows mapping a new patient into these analyses and predicting a personalized variable portrait. Specifically, the tool complements the SVM, 3-NN, and 9-NN classification probabilities by localizing the new patient in the 295-patients map and by providing a visual identity to compare with other patient and cluster portraits.
Application of the tool was iteratively improved with focus on impact of missing parameters and handling incomplete data as discussed above, and utilized to classify 14 patients newly recruited in the study.
.