J. Craig Venter

The human gut is inhabited by a complex and metabolically active microbial ecosystem regulating host health. While many studies have focused on the effect of individual microbial taxa, the metabolic potential of the entire gut microbial ecosystem has been largely under-explored. We characterised the gut microbiome of 1,004 twins via whole shotgun metagenomic sequencing (average 39M reads per sample). We observed greater similarity, across unrelated individuals, for functional metabolic pathways (82%) than for taxonomic composition (43%). We conducted a microbiota-wide association study linking both taxonomic information and microbial metabolic pathways with 673 blood and 713 faecal metabolites (Metabolon, Inc.). Metabolic pathways associated with 34% of blood and 95% of faecal metabolites, with over 18,000 significant associations, while species-level results identified less than 3,000 associations, suggesting that coordinated action of multiple taxa is required to affect the metabolome. Finally, we estimated that the microbiome mediated a crosstalk between 71% of faecal and 15% of blood metabolites, highlighting six key species (unclassified Subdoligranulum spp., Faecalibacterium prausnitzii, Roseburia inulinivorans, Methanobrevibacter smithii, Eubacterium rectale, and Akkermansia muciniphila). Because of the large inter-person variability in microbiome composition, our results underline the importance of studying gut microbial metabolic pathways rather than focusing purely on taxonomy to find therapeutic and diagnostic targets.

We report the results of a three-year precision medicine study that enrolled 1190 presumed healthy participants at a single research clinic. To enable a better assessment of disease risk and improve diagnosis, a precision health platform that integrates non-invasive functional measurements and clinical tests combined with whole genome sequencing (WGS) was developed. The platform included WGS, comprehensive quantitative non-contrast whole body (WB) and brain magnetic resonance imaging/angiography (MRI/MRA), computed tomography (CT) coronary artery calcium scoring, electrocardiogram, echocardiogram, continuous cardiac monitoring, clinical laboratory tests, and metabolomics. In our cohort, 24.3% had medically significant genetic findings (MSF) which may contribute to increased risk of disease. A total of 206 unique medically significant variants in 111 genes were identified, and forty individuals (3.4%) had more than one MSF. Phenotypic testing revealed: 34.2% of our cohort had a metabolomics profile suggestive of insulin resistance, 29.2% had elevated liver fat identified by MRI, 16.4% had clinically important cardiac structure or cardiac function abnormalities on cardiac MRI or ECHO, 8.8% had a high cardiovascular risk on CT coronary artery calcium scoring (Agatston calcium score > 400, Relative Risk of 7.2), 8.0% had arrhythmia found on continuous rhythm monitoring, 6.5% had cardiac conduction disorders found on EKG, 2% had previously undetected tumors detected by WB MRI, and 2.5% had previously undetected aneurysms detected by non-contrast MRI/MRA. Using family histories, personal histories, and test results, clinical and phenotypic findings were correlated with genomic findings in 130 study participants (63.1%) with high to moderate penetrance variants, suggesting the precision health platform improves the diagnostic process in asymptomatic individuals who were at risk. Cardiovascular and endocrine diseases achieved considerable clinical associations between MSFs and clinical phenotypes (89% and 72%, respectively). These findings demonstrate the value of integrating WGS and noninvasive clinical assessments for a rapid and integrated point-of-care clinical diagnosis of age-related diseases that contribute to premature mortality.

Modern medicine is rapidly moving towards a data-driven paradigm based on comprehensive multimodal health assessments. We collected 1,385 data features from diverse modalities, including metabolome, microbiome, genetics and advanced imaging, from 1,253 individuals and from a longitudinal validation cohort of 1,083 individuals. We utilized an ensemble of unsupervised machine learning techniques to identify multimodal biomarker signatures of health and disease risk. In particular, our method identified a set of cardiometabolic biomarkers that goes beyond standard clinical biomarkers, which were used to cluster individuals into distinct health profiles. Cluster membership was a better predictor for diabetes than established clinical biomarkers such as glucose, insulin resistance, and BMI. The novel biomarkers in the diabetes signature included 1-stearoyl-2-dihomo-linolenoyl-GPC and 1-(1-enyl-palmitoyl)-2-oleoyl-GPC. Another metabolite, cinnamoylglycine, was identified as a potential biomarker for both gut microbiome health and lean mass percentage. We also identified an early disease signature for hypertension, and individuals at-risk for a poor metabolic health outcome. We found novel associations between an uremic toxin, p-cresol sulfate, and the abundance of the microbiome genera Intestinimonas and an unclassified genus in the Erysipelotrichaceae family. Our methodology and results demonstrate the potential of multimodal data integration, from the identification of novel biomarker signatures to a data-driven stratification of individuals into disease subtypes and stages -- an essential step towards personalized, preventative health risk assessment.

In a recently published PNAS article, we studied the identifiability of genomic samples using machine learning methods [Lippert et al., 2017]. In a response, Erlich [2017] argued that our work contained major flaws. The main technical critique of Erlich [2017] builds on a simulation experiment that shows that our proposed algorithm, which uses only a genomic sample for identification, performed no better than a strategy that uses demographic variables. Below, we show why this comparison is misleading and provide a detailed discussion of the key critical points in our analyses that have been brought up in Erlich [2017] and in the media. Further, not only faces may be derived from DNA, but a wide range of phenotypes and demographic variables. In this light, the main contribution of Lippert et al. [2017] is an algorithm that identifies genomes of individuals by combining multiple DNA-based predictive models for a myriad of traits.

Significance: By associating deidentified genomic data with phenotypic measurements of the contributor, this work challenges current conceptions of genomic privacy. It has significant ethical and legal implications on personal privacy, the adequacy of informed consent, the viability and value of deidentification of data, the potential for police profiling, and more. We invite commentary and deliberation on the implications of these findings for research in genomics, investigatory practices, and the broader legal and ethical implications for society. Although some scholars and commentators have addressed the implications of DNA phenotyping, this work suggests that a deeper analysis is warranted.