From proteomics to personalized medicine: the road ahead

KEYWORDS:

• Proteomics
• omics pipeline
• mass spectrometry
• personalised medicine
• big data

The omics platform

Over the past decade, improvements in instrument sensitivity, speed, accuracy, and throughput, coupled with the development of technologies such as multiple reaction monitoring (MRM), Stable Isotope Standard Capture with Anti-Peptide Antibodies, Sequential Window Acquisition of all Theoretical Mass Spectra, cross-linking mass spectrometry, imaging mass spectrometry, imaging flow cytometry, and middle/top down proteomics, have led to significant advances in the field of proteomics [1–5]. Under the guidance of the Human Proteome Organisation (HUPO) over 80% of the proteins predicted by the human genome have now been identified using either mass spectrometric or antibody-based techniques, and the remaining ‘missing proteins’ are being steadily accounted for [6,7]. Resources such as the Human MRM Atlas, a comprehensive resource designed to enable scientists to perform quantitative analysis of all human proteins, are being developed to facilitate reproducible transfer of quantitative assays between laboratories. Such developments and initiatives now enable both in-depth discovery and targeted/quantitative workflows, opening the door to the clinical diagnostic arena. Coupled with this, the establishment of comprehensive databases and the development of powerful in silico techniques is enabling effective data mining. In particular this has enabled interactome studies allowing the identification of key signaling pathways leading to potential new drug targets, although to date it has been estimated that less than 20% of the protein interactions in humans, not counting dynamic, tissue- or disease-specific interactions, have been identified [8].

Personalized medicine

The emergence of the ‘omics’ platforms (genomics, proteomics, metabolomics, transcriptomics, and interactomics) now gives us a pipeline around which to develop the infrastructure required for personalized (precision, P4) medicine. While the concept of personalized medicine is not new (it was actually described in The Yellow Emperors Canon of Internal Medicine over 2000 years ago [9]), the concept has gained momentum since the turn of this century. At the proteomics level this concept has been championed by Prof Lee Hood at the Institute of Systems Biology in Seattle who has coined the phrase P4 medicine: predictive, preventive, personalized and participatory [10]. He realized that a multidisciplinary systems biology approach was required, combining the concerted effort of specialists from a wide variety of disciplines (e.g. medicine, chemistry, biochemistry, physics, mathematics, computing, bioinformatics, and manufacturing), to bring together the multiple skills and technologies required.

Personalized medicine involves specifically tailoring treatment to the individual characteristics of the patient rather than the current approach of stratifying patients into treatment groups based on phenotype. It will address both health and disease and impact on predisposition, screening, diagnosis, prognosis, pharmacogenomics, and surveillance [11], based on a comprehensive understanding of an individual’s own biology. Importantly the patient now defines his own normal levels, facilitating the correct interpretation of biomarker assays. Personalized medicine is predicted to significantly reduce global health budgets, both by reducing the need for hospitalization and associated costly procedures, and by minimizing the unnecessary/inappropriate use of drugs (the top ten highest-grossing drugs in the United States only help from 4% to 25% of the patients who take them) [12].

There is no doubt that a personalized approach to medicine is viable and effective as evidenced by the recent study by Chen et al. [13] in which a comprehensive omics characterization (Integrative Personal Omics Profile (iPOP), an analysis that combines genomic, transcriptomic, proteomic, metabolomic, and autoantibody profiles) of an individual over a 14-month period revealed dynamic molecular and medical phenotypes including a risk for type 2 diabetes. This study alone generated more than 30 terabytes of data [14] highlighting the massive computing support that will be required for personalized medicine.

The Big Data problem

One of the major hurdles in realizing the goal of personalized medicine, and currently perhaps the major bottleneck, will be effectively handling and storing the enormous amounts of data that will be generated (the big data problem). It has been estimated that the size of ‘Big Data’ in the US health-care system alone was around 150 exabytes in 2011, is increasing at a rate of between 1.2 and 2.4 exabytes per year, and will rise faster as additional new data intensive technologies are adopted (e.g. microarrays, whole genome sequencing, and imaging). This can be extrapolated to suggest that, before long, countries with large populations and emerging economies, such as China and India, will be generating zettabyte (10²¹) to yottabyte (10²⁴) amounts of health-related data each year [15], although the actual levels that will need to be stored will depend on how the data is finally summarized and processed.

Big Data has 6 key considerations, known as the 6 Vs: value, volume, velocity, variety, veracity, and variability [15] that must be taken into account when designing suitable solutions to ‘Big Data’ analysis. Consideration must include: What information is needed? How to handle the complexity of information? How to filter unnecessary data? How to integrate the data in various formats from very heterogeneous resources? What kind of principles or standards need be adopted? How will data integrity be monitored?

Clearly a first requisite is rapid and facile data deposition and retrieval. However, even for the raw MS data this has not proved trivial, and there has been a relatively poor uptake of this essential action by the proteomics community, possibly due to burdensome and unreliable data deposition and retrieval protocols used in some earlier attempts. This has now been largely resolved, and, importantly, a number of the proteomics journals are now making data deposition mandatory for publication.

Robust tools will be required for depositing and accessing all the necessary data. Coupled with this, robust storage with automated backup is essential. To ensure uniformity of data, Standard Operating Procedures across platforms are necessary and general acceptance of best practices will be required [16]. Acceptance of a ‘common language’ between both vendors and scientific communities will facilitate this. Cloud computing will provide rapid, cost-effective analysis of the enormous amounts of data generated. Interestingly, at ACMS 2015 two companies (SCIEX and Thermofisher) launched cloud-based applications aimed at integrating multiple platforms.

Ethical concerns

Another major hurdle that needs to be addressed sooner rather than later is that of ethics. Numerous concerns have been expressed in this area [17], the most important probably being those of informed consent, privacy, ownership, and the difficulty of providing meaningful access rights to individual data subjects who lack the necessary resources. It is important to tackle this issue globally with all key parties having a seat at the table.

The funding dilemma

How will funding for such enormous projects be addressed? It has been pointed out recently [18] that this will almost certainly face some difficulties, particularly with grants involving multidisciplinary, multi-institutional, and often multinational groups of the type required to support personalized medicine. For the proteomics community funding may face additional hurdles. Poste has noted that the current herd mentality that portrays genome sequencing as a panacea for understanding all aspects of disease pathogenesis is unrealistic. This was recently further reinforced in a presentation by Prof Ralph Bradshaw at the recent ‘The Omics Revolution: Uncovering the Complexity of the Human Proteome’ meeting held in Kingscliff, Australia (October 2015), where he proposed that genomics and metabolomics were currently ‘picking the low hanging fruit’ and monopolizing funding, possibly because the type of data they produce and the analyses required are far simpler than those of most large-scale proteomic experiments. In this respect it is interesting to note that the recently announced Obama US Personalised Medicine Initiative, whereby US$ 215 million were injected into the health care system to monitor a research cohort of one million or more citizens, made no mention of proteomics in its initial announcements, although this ‘glaring omission’ was later corrected.

The road ahead

Clearly the introduction of personalized medicine holds enormous promise and benefits for mankind. It has already established a foothold, with more than 25% of all new drugs approved by the US FDA in 2015 relating to personalized medicine [19]. However, we are currently only looking at the tip of the iceberg. The switch toward personalized medicine will require many changes in the way we think and the manner in which things are done. Aside from the hurdles mentioned above (the size of the data, ethical, and funding issues) suitable reimbursement policies will need to be introduced to cover the new technologies, drugs, and services to ensure patients have full access to personalized medicine. Alongside this, the education system must be updated to prepare the next generation of doctors and other health care professionals for the road ahead. Now is the time to ensure that all aspects of this paradigm shift in health care are carefully thought through and the correct global infrastructure developed.

Financial and competing interests disclosure

The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.