Biomarker discovery in colon cancer through next-generation metabolomics

About This Grant

PROJECT SUMMARY Metabolomics is the study of small molecule metabolites and has drastically improved health research, diagnostics, and patient treatment. Metabolomic approaches quantify the abundance of small molecules within biological samples using liquid chromatography coupled to mass spectrometry (LC/MS), which is capable of simultaneous detection of 1,000s of metabolites. Cancer, a fundamentally metabolic disease, has benefitted greatly from metabolomics, having enabled the discovery of tumor heterogeneity, novel biomarkers, and individualized treatments. Until recently, however, nearly all large-scale metabolomics analyses have been limited to targeted studies of a few hundred pre-discovered metabolites. In part, this is due to the data burden and uncertainty in global analysis of all detected metabolites (i.e., untargeted metabolomics). Researchers have been slow to adopt untargeted metabolomics because it relies heavily on software to evaluate the large number of signals detected. Unfortunately, current software are limited by two major factors: (1) ~90% of the detected signals do not represent unique metabolites but are false positives or redundant compound detections derived from chemical or electronic noise that current software does not address, and (2) More than 80% of measured metabolites are removed from downstream analysis due to ambiguous identifications, in large part due to software programs not providing statistical scores for metabolite identification certainty. These limitations lead to an inflated data burden, poor metabolome coverage, and misidentified metabolites, diminishing the value and impact of untargeted metabolomics to cancer research. At Panome Bio, we are the first metabolomics CRO to bring a truly untargeted metabolomics approach to biomedical researchers. Our company currently offers untargeted metabolomics in a fee-for-service model and has achieved strong initial traction, completing >100 customer projects to date. An important aspect to this platform is MassID, our Next-Generation untargeted metabolomics analysis software. While effective, our current version of MassID (MassID 1.0) is designed to focus only on the most confidently identified metabolites at the expense of removing ambiguous (but informative) data. Here, we propose to develop two software components that will improve the scoring and filtering systems used during metabolite identification: PeakDefrag and DecoID2. These will be integrated into a second-generation analysis software (MassID 2.0) that is hosted on the cloud and available through a web application, offering all researchers straightforward access to their data. The grant will be completed by benchmarking MassID 2.0 against MassID 1.0 and other untargeted software approaches in a colon cancer cohort. With these results, we plan to be the first metabolomics company to back metabolite identifications with statistical values and to further our position as the leading platform in eliminating the number of signals derived from noise. We expect Mass ID 2.0 to transform untargeted methods by increasing the number of statistically significant metabolite identifications and to advance Panome Bio as an industry leader in multi-omics discovery research, a market valued at $750M.