The publication BioCentury Innovation recently featured the ongoing GNS Healthcare collaboration with the Multiple Myeloma Research Foundation (MMRF), highlighting the project’s pioneering use of causal analysis to accelerate the discovery of innovative treatments for patients with multiple myeloma.
“I had this conviction that if we had enough data from enough patients with various interventions, including the outcomes, that we would be able to turn that data into computer models” that could help select the best treatments for patients, and identify new ones, GNS CEO Colin Hill told BioCentury.
In “Cause, Not Correlation,” Hill explained that the unprecedented size of the MMRF dataset enables use of causal analysis for biomedical discovery by GNS and that, while some academics conduct causal analyses using big data, applying causal analysis for biomedical discovery is groundbreaking.
The MMRF-GNS project leverages the GNS MAX™ architecture and patented REFS™ inference engine and simulation platforms to analyze data collected as part of the landmark MMRF CoMMpass study. The CoMMpass dataset delivers on GNS’s vision by delivering a large volume of multi-layered data, including DNA sequence variation, gene expression, and clinical outcomes. The longitudinal study of 1,000 newly diagnosed patients with active multiple myeloma represents the largest multiple myeloma patient registry.
The GNS machine learning engine, REFS, is being used to identify causal drivers and underlying molecular processes of disease progression. REFS will discover the most likely targets for therapeutics to treat, and perhaps to predict and prevent, relapses and refractory disease. In addition, REFS will reveal predictive diagnostic biomarkers that determine which treatments will work and for which patients. The work promises to support physicians and patients with a broader set of individualized treatment protocols mapped to specific molecular and genomic characteristics.
GNS and MMRF hope to publish initial findings by the end of 2015.