Drug target identification
Patient stratification for improved diagnostic and treatment
Metabolomics is the study of chemical processes involving metabolites, and integrative analysis of metabolites can provide a comprehensive view of dysregulated biological pathways leading to a disease. Liquid chromatography mass spectrometry (LC/MS) can compare levels of metabolites between disease and control systems; however, identification of each LC/MS peak requires an additional experiment, such as tandem MS, to identify each LC/MS peak. Moreover, only a fraction of LC/MS peak features can be identified using tandem MS, making the analysis low-throughput, costly and time consuming.
This invention presents a novel systems biology approach for more efficiently analyzing untargeted metabolomics data to identify disease-modifying metabolites that remained undetected experimentally. This novel network-based approach infers dysregulated pathways and components from the differential metabolite features between control and disease system. This invention effectively eliminates a bottleneck problem in metabolomics studies with LC/MS by reducing the need for additional experiments for metabolic identification.
As a proof of concept, the inventors have applied this novel algorithm to analyze untargeted lipid profiling data from a Huntington’s disease model. They identified and experimentally validated a suspect network with altered levels of metabolites and phosphoproteins whose components can be considered drug targets.
Enables high-throughput metabolomics data analysis.
Able to predict the mechanisms causing metabolite dysregulation.
Allows integrative analysis of untargeted metabolomics with various high-throughput experimental data, such as proteomic sequencing data.
Reduces the need for additional experiments for metabolic identification.