by Mike Milburn

When most people think of precision medicine, genomics is top of mind. While genomics will play an undeniably important role in realizing precision medicine, most researchers recognize the need to include other types of phenotypic data in addition to medical history and standard clinical assessment. 

The last decade of genomics research has revealed a higher-than-anticipated individual genetic variation; that most traits of interest involve a combination of many genes; and, the majority of mutations of interest reside in non-coding regions of the genome, where we have a very poor understanding of the function. Influences from the microbiome, epigenetics and the environment add significant complexities. In fact, getting your whole genome sequenced provides a great deal of data, but less-than-desired tangible health information.  

Basically, the vast majority of diseases don’t have a known gene association. There is growing appreciation that a combination of genetic and non-genetic factors causes complex illnesses such as diabetes, cancer, cardiovascular and neurological diseases. New approaches to boost the success rate and identification of disease-causing genes are essential. Clinicians must take into account the impact of these factors to make an informed diagnosis and apply precision medicine. 

Last week, PNAS published an important paper for Metabolon. We conducted a study with Dr. Tom Caskey at Baylor College of Medicine and 80 healthy adults from his medical practice. There were 45 men and 35 women, with an average age of 54 years. Each volunteer provided a detailed medical history, and whole-exome sequencing was obtained. None of the volunteers reported any serious diseases. Healthy, right? 

Our global metabolomics technology works much like your typical physician-ordered blood test, but rather than measuring a handful of things in the blood, we measure upwards of 500-700 things. After running these 80 healthy blood samples on our platform, we profiled 575 metabolites covering 72 biochemical pathways.      

Surprisingly, we identified tentative medical findings in nearly 25 percent of the volunteers. These ranged from potential damaging genetic mutations that were previously unidentified to early signs of diabetes, liver dysfunction and gut microbiome problems. We also observed metabolic signatures associated with potential drug toxicity effects. 

This is why a growing number of large precision medicine and next-generation sequencing (NGS) initiatives have adopted metabolomics as a cornerstone of their programs to link genetics and metabolic profiles to phenotypes or health states. Metabolomics reflects the influences of genes, diet, lifestyle and environment to aid in understanding gene function and how diseases originate, and it could provide the biomarkers for health assessment and customized drug therapy.  

Large population studies, such as genome-wide association analyses, have shown that combining metabolomics with genomics is a valuable approach to gain new understanding of genetic variance and disease risk. In this study, we integrated genomic and metabolomic data in an attempt to improve medical interpretation of an individual’s disease risk in a small clinical cohort. 

To our knowledge, our study is the first to apply non-targeted metabolomics with NGS for individual clinical assessment. The results revealed that metabolomics can give significant insights with clinical importance to health assessment and disease management in our goal toward precision medicine.