The Age of Omics-Driven Precision Medicine.

Abstract

The rapid evolution of omics technologies has the unprecedented potential to unravel the mysteries underlying human health. Omics, which describe the collective technologies used to explore entire sets of genes, RNA, proteins, or molecules in biological samples, are powerful new tools used to scrutinize physiological and environmental processes responsible for disease and therapeutic responses. Integration of the knowledge gained from these technologies will ultimately advance our efforts to prevent, detect, and treat many diseases. The first omics technology, genomics, appeared in the literature in 1987.1 However, approximately 15 years elapsed until publication of the full human genome sequence and the beginning of the so-called “post-genomic era,” which inspired the development of other omics technologies.2 Knowledge of the human genome facilitated rapid determination of the gene sequences of proteins and other macromolecules of interest. This groundbreaking information spurred researchers to move beyond the genome and begin to study the biological roles of gene products and their implications in disease.2 Continued research has produced new omics technologies, including transcriptomics, proteomics, and metabolomics, which, along with advances in bioinformatics and computational approaches, enable detailed investigation of complex biological processes with greater sensitivity and resolution compared with traditional biochemical methods. Whereas traditional methods are often time-consuming and target specific, omics technologies offer high-throughput, unbiased approaches for investigating multiple targets simultaneously in a relatively time-efficient manner. Due to these advantages, omics technologies have become powerful additions to our arsenal of tools used in the continuum of drug discovery, development, and utilization. With computational and analytical advances, omics continue to expand the breadth and complexity of methodologies at our disposal, allowing researchers and clinicians to delineate underlying mechanisms of disease and variability in drug response, which form the foundation of precision medicine. Genomics, which refers to the study of genetic composition, was originally coined to indicate analysis of data generated from DNA sequencing projects.1 However, due to technological advances, applications of genomics have rapidly evolved and now extend to personalized medicine via pharmacogenomics. Pharmacogenomics is the study of how interindividual differences in gene sequences, such as single-nucleotide polymorphisms (SNPs), can alter an individual's response to drugs.3 SNPs that influence the absorption, distribution, and clearance of drugs, including those in genes encoding drug metabolizing enzymes and transporters, can alter an individual's exposure to drugs. Numerous SNPs have been linked to altered blood/tissue drug concentrations and increased adverse drug effects.4 Likewise, SNPs can alter either exposure or binding to drug receptors and other targets, thus impacting therapeutic effects. Personalized medicine aims to use genomics to identify variations in a patient's genome that may inform their risk of adverse effects or therapeutic failure, allowing healthcare providers to select the optimal medication and dose for a given patient. Initiatives such as eMERGE-PGx, which used pharmacogenomics to examine sequence variations in 82 pharmacogenes in more than 5,000 individuals, are contributing valuable insight into mechanisms underlying individual variability in drug response and potential implications.5 However, further research is needed for the widespread clinical application of pharmacogenomics. A recent review of clinical pharmacogenomic implementation studies in the United States6 provided recommendations for effective implementation, including the need to educate community-based practitioners and pharmacists to enable broad and comprehensive pharmacogenomic services. The evolution of genomics has begun to encompass not only an individual's genetic information but also genetic information on the diverse microorganisms that reside on or within their person or in the surrounding environment. This exciting application of genomics, termed metagenomics, examines genetic material contained within the collections of these microorganisms.7 Metagenomics has emerged as a promising technique to identify the microbial species contained within a complex environmental or biological matrix and characterize their genetic makeup.8 In contrast to traditional methods used to identify or sequence the genes of microorganisms, metagenomics obviates having to culture each individual organism, which is not feasible for many species. Applications of metagenomics could include characterizing the human gut microbiota, as imbalances in gut flora have been implicated in a number of diseases, such as obesity and diabetes,7 or identifying genes involved in antibiotic resistance (the “resistome”).8 As detailed in the State of the Art review by Sukhum and colleagues,9 metagenomic approaches involving high-throughput DNA sequencing and classical culture-based assays are being used to identify, characterize, and track emerging and antibiotic-resistant pathogen outbreaks across patients and environmental reservoirs. Information generated from these integrated analyses, as well as identification of high-risk antibiotic resistance genes, has provided new insights into resistance and may be instrumental in the development of novel therapeutic agents that can defeat drug-resistant pathogens. Proteomics, which refers to the analysis of proteins, has become increasingly more accessible and successful with improvements in chromatography and mass spectrometry analysis techniques. Proteomics can identify and quantify not only a large percentage of proteins present in complex biological samples but also post-translational modifications (PTMs) such as phosphorylation and glycosylation.1 Identification of PTMs is of increasing interest, as aberrant PTMs are believed to be involved in the pathogenesis of several chronic illnesses. The white paper by Prasad and colleagues10 highlights this capability of proteomics, as well as details the need for harmonized guidelines regarding applications of proteomics to translational pharmacology. A subtype of proteomics, chemoproteomics, is arguably one of the most promising applications of omics to the field of drug discovery. Functional small molecules labeled with analytically detectable probes are exposed to the entire proteome, and target proteins bound to the small molecules can be detected and identified using high-throughput mass spectrometry techniques.2, 11 By generating small molecules that recognize and/or inhibit large protein classes, high-throughput chemoproteomics is anticipated to identify novel druggable targets previously overlooked during drug discovery.11 Using small molecules that inhibit malarial cysteine proteases, enzyme targets that impair the parasite's life cycle have been identified and have led to the development of novel antimalarial agents.11 Another noteworthy application of chemoproteomics could involve identification of small molecules that target nuclear hormone receptors, which contribute to the regulation of drug metabolizing enzyme and transporter expression. Another application of proteomics to biomedical research is toxicoproteomics. As the term implies, toxicoproteomics refers to the use of proteomics to assess toxicant-induced changes in protein expression.2 Toxicoproteomics is currently used to study mechanisms of drug-induced organ toxicity, with one ultimate goal being the discovery of biomarkers that are detectable in saliva, blood, or urine samples. In line with this goal, toxicoproteomic approaches have been applied to a number of drugs known to induce toxicity in the liver, kidney, brain, and lungs.12 Although other omics strategies, including transcriptomics, have been used to assess changes in gene expression following drug-induced toxicity,2 toxicoproteomic approaches can detect toxicant-induced changes in PTMs, which could provide further insight into mechanisms of drug-induced toxicity. For example, PTMs caused by inflammation, oxidative stress, and carbamylation have been implicated in numerous chronic diseases, including cancer, chronic renal failure, and rheumatoid arthritis.12 Thus, toxicoproteomics, which can detect changes in both protein expression and PTMs, provides a powerful tool for investigating mechanisms of drug-induced toxicity. Endogenous metabolites encompasses all small molecules produced as biological endpoints of cellular processes and are subject to alterations due to genetic, pharmaceutical, or environmental factors.2, 13 Moreover, disease initiation and progression are often caused by or associated with dysfunctional internal metabolic processes. Metabolomics allows us to detect disturbances in circulating metabolites in biological samples such as blood, urine, feces, and saliva.2, 13, 14 In contrast to gene or protein sequences, endogenous metabolites have the same chemical structures across species, thus increasing the translational power of these biomarkers with respect to drug discovery and development.13 Pharmaceutical interventions can cause intentional or unintentional metabolic changes associated with response to therapy and/or treatment-associated adverse effects.2, 13 Metabolic enzymes can activate or deactivate drugs, and drugs and their metabolites can alter homeostatic concentrations of endogenous metabolites in circulation.2 Pharmacometabolomics, which specifically studies alterations due to pharmaceutical interventions, can be used to evaluate drug toxicity, therapeutic response, and mechanism of action.2 Hence, pharmacometabolomics can be used to predict how a patient with a specific metabolic profile may respond to a drug, determine off-target effects, and identify early predictors of therapeutic efficacy.2, 13 The State of the Art review by Hu and colleagues describes a number of specific examples of which pharmacometabolomics was used to identify diagnostic and prognostic biomarkers of various diseases, elucidate disease mechanisms, identify novel drug targets, and predict drug response.15 Due to the high-throughput nature of omics technologies, analysis of the large data sets generated requires strategic data science management.2 The complexity of these data sets increases the probability of obtaining false positives or false negatives.14 Current analytical methods are discordant in terms of their ability to produce true positive (sensitivity) and true negative (specificity) results.2 Because consistency and standard practices across studies are lacking, interpretation of omics output can be challenging; controversy exists regarding choice of biological samples, how samples are collected and stored, and how data are collected and analyzed.2, 13, 14 Additionally, biological molecules such as genes, proteins, and metabolites are dynamic. Even for the same individual under the same analytical conditions, the data will be analyzed as static and may not be representative of the dynamic nature of biological processes.2 Thus, unbiased interpretation of the data can be problematic. Indeed, validation of various disease biomarkers is one of the major challenges in omics research.2 A second limitation to omics approaches is that current analytical technologies can detect only a fraction of the thousands of genes, proteins, and metabolites in humans.2 In addition, incremental changes in biomarker concentrations that could indicate initiation of pathological processes often cannot be detected.2, 14 The inherent variability in humans with respect to age, sex, environment, lifestyle, and/or diet further complicates the practical utility of omics approaches.13, 14 Such heterogeneity can mask or alter the omics profile of an individual, making identification of disease processes difficult.13 Interpretation of the data from one omics approach is already challenging. Thus, combining data from multiple types of applications renders interpretation exceedingly challenging and prone to bias.2 A potential solution is to build network maps for biological processes based on the data obtained from multiple omics studies, which rely on the accuracy of the data science management.2 However, the complete mechanism of action of many drugs is unknown for various pathologies, making these network maps difficult.13 As Fohler and colleagues16 note in their macroscopy article, the knowledge generated from innovative omics research has the potential to optimize therapeutics and human health; however, effective translation into clinical practice is required. Clinical pharmacologists play crucial roles in not only the drug discovery and development continuum but also the education and engagement of patients and communities, which are essential for knowledge translation. Therefore, clinical pharmacologists are best suited to transform novel advances to patient care. Hence, interdisciplinary teams involving clinical pharmacologists can work collectively toward integration of patient omics profiles to guide personalized care. This strategy encompasses all aspects of disease, from diagnosis at the bench to bedside therapeutics. In the future, omics could enable clinical pharmacologists and healthcare providers to develop tools and strategies for disease prevention. The “All of Us” program, funded by the National Institutes of Health and involving many clinical pharmacology research sites, is a transparent research initiative that aims to collect information from over 1 million individuals in the United States and uses multiple omics technologies to determine factors that contribute to disease.17 This program will allow researchers to identify biological markers that signal risk or underlying cause of disease, including factors that determine individual response to drugs, all while empowering participants with information they can use to improve their own health. This program will also identify causes of intersubject variability, disease susceptibility, and drug responses by taking into account biology, environment, and lifestyle choices (Figure 1). Developing a comprehensive database of the whole body, rather than organ-specific function of drugs would help advance the clinical applications of omics technologies. Omics approaches can be applied during the entire drug-development spectrum, from target discovery to personalized medicine. These approaches can improve the drug-development process and aid in the prediction and evaluation of efficacy and adverse responses. Omics can also help optimize existing treatments and identify safer, more effective therapies. Further application to network biology would create a useful tool to demystify the incredibly complex pathophysiology of disease and identify new therapeutic targets. Theoretically, integrated omics would permit detection of the slightest changes in biological activity and enable discovery of clinically relevant biomarkers. However, before application in the clinic, improvements in bioinformatics processing of large data sets is required. Future studies should focus on developing, validating, and standardizing analytical platforms and computational methods to interpret the large data sets without bias. Next steps include validating the analytes obtained in large population-based studies. Such strategies would enhance reproducibility and reliability of studies conducted in laboratories worldwide. Armed with a clear understanding of the omics and their contribution to health and disease, futuristic health care may truly achieve its potential role in precision medicine. The authors have no conflicts of interest to disclose.