https://orcid.org/0000-0003-0474-8668
1. Introduction
The last few years have witnessed the emergence of personalized precision medicine (PPM), a novel approach to improve clinical decisions, integrating individual multi-omic profiles, and heterogeneous sources of information of the electronic health record (EHR) including medical imaging and the large amount of data generated through electronic devices that could reflect the course of the disease, lifestyle, and environment [Citation1]. PPM is expected to contribute both socially and economically by providing cost-effective tools for the prevention, diagnosis, and treatment of diseases with direct benefits to the health and quality of life of the population.
However, PPM faces practical questions and challenges in its clinical implementation that involves economical, technical, and ethical considerations. One of the most critical challenges are how to manage, analyze and interpret very large, complex, and heterogeneous data preserving security and privacy requirements of medical information. Bioinformatics aims to directly tackle this challenge exploiting data availability with new technologies and computational infrastructures to produce PPM approaches and solutions for clinical decision-making.
2. The shift in the paradigm: from evidence-based to data-driven medicine
Evidence-based medicine (EBM) originated in the early 1990s to educate clinicians in the interpretation of the rapidly accumulating clinical research. EBM proposes a hierarchical classification of clinical research in terms of quality and robustness of experimental and statistical design, enabling clinicians to develop solid, evidence-based clinical practise guidelines. A crucial limitation of EBM is that it focuses on population-based trials which fail to fully recapitulate interpatient heterogeneity, which plays a crucial role in diseases such as cancer and rare congenital disorders. Interestingly, PPM data driven approach is better suited to address these limitations.
PPM is situated in the paradigm shift from evidence-based to data-based medicine, with data analysis being an effective way of addressing well-formulated clinical questions for clinical decision-making. Current examples of PPM application to clinical practice include hereditary disorders and cancer. Hereditary disorders such as congenital hearing loss have greatly benefited from advances in genetic testing, providing an earlier and more accurate diagnosis and genetic counseling [Citation2]. In cancer, the current clinical routine includes genetic and molecular biomarkers for diagnosis and to improve patient stratification and prognosis (i.e. KRAS mutations in anti-EGFR target therapy or PDL1 expression in immunotherapy response). Nevertheless, the existence of non-responders’ individuals still indicates that these single biomarkers are not sufficient to establish an accurate patient selection [Citation3]. Thus, cancer PPM faces important challenges like inter-patient and intra-tumoral heterogeneity which have a crucial role in cancer patients’ response to therapy and in the development of drug resistance [Citation4,Citation5].
Current sequencing technologies and the rapid increase of real-time biomedical monitoring instruments, biosensors, and wearable devices have allowed researchers to retrieve multi-omics, epidemiological, health care, and wellness data on an unprecedented scale. Such data combined with EHR information will be the key to perform a better patient stratification that entails more accurate diagnosis and treatments [Citation1]. Moreover, multi-source data integration may also reduce side effects originated by some current therapeutic strategies (i.e. chemotherapy), providing a more accurate assessment of disease evolution and prognosis. Accordingly, it is expected that these improvements may also, in turn, alleviate the economic burden associated with inaccurate patient diagnosis and therapy side effects [Citation6].
In light of the possibilities of PPM several national and international efforts and consortiums have been launched to bring this clinical paradigm closer to the patients involving efforts to drive multi-source data sharing.
Some remarkable examples include: the ‘All of Us’ Research Program (https://www.researchallofus.org/), the GenomeAsia100 k (https://genomeasia100k.org/), the 100.000 Genomes Project (https://www.genomicsengland.co.uk/), the European 1 Million Genomes Initiative (https://ec.europa.eu/digital-single-market/en/european-1-million-genomes-initiative), the ICGC-Argo (https://www.icgc-argo.org/) and the ICPerMed consortium (https://www.icpermed.eu/).
It is important to highlight the key role of data sharing initiatives for PPM implementation. For instance, cancer PPM is currently hampered by the lack of complete clinical and histopathological records in the cancer genomics public databases. In this context, data sharing arises as a rational alternative toward more comprehensive data sources and subsequently more accurate tools. In addition, the possibility of expanding the availability of health-related data through extensive digitization of medical records and collaboration among countries for the secure exchange of such data will represent a paradigm shift not only for policymakers and biomedical scientists, but also for citizens. Accordingly, a number of regulatory frameworks and proposals to establish legal margins to access human genetic and phenotypic data avoiding health inequities in PPM-based initiatives are taking their first steps [Citation7]. In this sense, the establishment of solidarity-based legal frameworks will be also key in data sharing across health care systems while fundamental rights are respected and data privacy and security are guaranteed [Citation8].
3. Precision bioinformatics, the key to personalized precision medicine
The successful implementation of PPM implies to afford several actual challenges that encompass: i) the incorporation of infrastructures, technologies, services, tools, and applications, capable of handling big data from multiple sources ii) to associate molecular and clinical heterogeneous information to facilitate clinical decision-making iii) to protect, securely share and maintain patients’ data and iv) to recruit qualified professionals able to interrogate, analyze and interpret this information in the healthcare system. These challenges are being faced by ‘precision bioinformatics’ defined as the branch of translational bioinformatics specialized in the development of computational infrastructures, methodologies, and tools required for clinical studies adopting the PPM [Citation9].
Current European computational infrastructures are well-positioned to support precision bioinformatics. Several projects and initiatives are actively participating in planning the future of HPC, biomedical software, and the implementation of widespread genomic data access. For instance, ELIXIR intergovernmental organization provides a stable and sustainable infrastructure of life science resources across Europe including databases, software tools, training materials, cloud storage, and supercomputing infrastructures able to manage heterogeneous big data [Citation10].
Data patient access has also the development of software solutions, analytical tools, and predictive models for PPM. Artificial Intelligence (AI) assisted algorithms to predict therapy response and clinical risk predictions or in silico drug prescription tools are examples of the required tools for PPM [Citation11]. AI-based tools have shown the capacity to obtain useful information from complex data. For instance, successful AI predictive models using medical imaging are already available to provide a more accurate way to predict disease progression [Citation12]. These tools are expected to grow exponentially while increasing accessibility to quality data and improving compatibility with robust technology platforms.
Overcoming PPM challenges involve many stakeholders and require coordinated cross-border initiatives for setting up bioinformatics infrastructures and technologies. PPM tools applicability is currently getting ready in academia, research institutions, and biotech companies being the final objective to integrate them in hospitals and health systems. Undoubtedly, professionals with a background in computer science, statistics, molecular biology, and clinical practise are a requisite to face all these challenges.
4. Clinical bioinformaticians: the new guys in the office
Emerging clinical bioinformatics groups with computational, biomedical, and clinical skills are being incorporated into hospitals to bridge bioinformatics, data science, and clinics. Similarly, to a number of medical specialties, such professionals should be integrated into health-care multidisciplinary teams together with other health experts (e.g. physicians, pharmacists, etc.). Clinical bioinformaticians are an asset in hospitals and actively interact as part of healthcare teams focusing on the issues in which they specialize to improve the care, diagnosis and treatment of patients. In this scenario, the interdependence between clinical bioinformaticians and other health professionals is key [Citation9].
The clinical bioinformaticians can empower the healthcare providers and play major independent roles. More specifically, clinical bioinformaticians tasks will cover the needs associated with the implementation of data-based medicine. These tasks include: i) to develop platforms for big data processing and maintain computational pipelines; ii) to perform data analysis on heterogeneous biomedical data sources; iii) to access and harness EHR information systematically; iv) smartly questioning the data to provide intuitive reports; v) to fluently communicate relevant findings to other health professionals to guide clinical decision-making. Carrying out such tasks requires trained experts such as IT specialists, software developers, data miners, AI professionals, big data scientists, and data interpreters (of course not in one single person).
Training of health professionals would help to minimize the gap between medical and data science knowledge [Citation13]. To bridge this gap, a number of universities have developed new postgraduate programs and international organizations like ELIXIR and GOBLET have launched educational initiatives for a broad spectrum of expertise levels such as the Train-the-Trainer workshops [Citation14]. These projects cover not only specific bioinformatics topics but also general principles crucial to provide robust training.
In conclusion, bioinformatics for routine application of PPM is getting ready. National and international roadmaps to provide sustainable standards-based infrastructures for data exchange, archiving and computing are in an advanced state of implementation [Citation15]. Efforts on the development of AI algorithms, machine learning techniques, and big data-based predictive models for PPM are increasingly demanded since they are critical to foster the switch from the evidence-based to the data-driven personalized precision medicine. Consequently, precision bioinformatics trained experts are being progressively recruited by healthcare institutions and biotechnological and pharmaceutical companies. All of the above relies on the availability of controlled-access human data, data sharing efficient models and solidarity perspective for further data uses that may improve healthcare in other individuals. Health care in the age of PPM will become a global health care and bioinformatics will provide the basis of this change.
Declaration of interest
The authors have no other 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 apart from those disclosed.
Reviewers Disclosure
Peer reviewers on this manuscript have no relevant financial relationships or otherwise to disclose.
Additional information
Funding
References
- Shameer K, Badgeley MA, Miotto R, et al. Translational bioinformatics in the era of real-time biomedical, health care and wellness data streams. Brief Bioinform. 2017 Jan;18(1):105–124.
- Rudman JR, Mei C, Bressler SE, et al. Precision medicine in hearing loss. J Genet Genomics. 2018 Feb 20;45(2):99–109.
- De Guillebon E, Dardenne A, Saldmann A, et al. Beyond the concept of cold and hot tumors for the development of novel predictive biomarkers and the rational design of immunotherapy combination. Int J Cancer. 2020 Jan 30. DOI:10.1002/ijc.32889
- Jamal-Hanjani M, Quezada SA, Larkin J, et al. Translational implications of tumor heterogeneity. Clin Cancer Res. 2015 Mar 15;21(6):1258–1266.
- Turajlic S, Sottoriva A, Graham T, et al. Resolving genetic heterogeneity in cancer. Nat Rev Genet. 2019 Jul;20(7):404–416.
- Gavan SP, Thompson AJ, Payne K. The economic case for precision medicine. Expert Rev Precis Med Drug Dev. 2018;3(1):1–9.
- Wolf SM, Bonham VL, Bruce MA. How can law support development of genomics and precision medicine to advance health equity and reduce disparities? Ethn Dis. 2019 Dec 12;29(Suppl 3):623–628.
- Prainsack B. The “We” in the “Me”: solidarity and health care in the era of personalized medicine. Sci Technol Hum Values. 2018;43(1):21–44.
- Gómez-López G, Dopazo J, Cigudosa JC, et al. Precision medicine needs pioneering clinical bioinformaticians. Brief Bioinform. 2019 May 21;20(3):752–766.
- Drysdale R, Cook CE, Petryszak R, et al. The ELIXIR core data resources: fundamental infrastructure for the life sciences. Bioinformatics. 2020 Jan 17. DOI:10.1093/bioinformatics/btz959
- Piñeiro-Yáñez E, Reboiro-Jato M, Gómez-López G, et al. PanDrugs: a novel method to prioritize anticancer drug treatments according to individual genomic data. Genome Med. 2018 May 31;10(1). DOI:10.1186/s13073-018-0546-1
- Baptista D, Ferreira PG, Rocha M. Deep learning for drug response prediction in cancer. Brief Bioinform. 2020 Jan 17. DOI:10.1093/bib/bbz171
- Attwood TK, Blackford S, Brazas MD, et al. A global perspective on evolving bioinformatics and data science training needs. Brief Bioinform. 2019 Mar;20(2):398–404.
- Via A, Attwood TK, Fernandes PL, et al. A new pan-European Train-the-Trainer programme for bioinformatics: pilot results on feasibility, utility and sustainability of learning. Brief Bioinform. 2019 Mar;20(2):405–415.
- Saunders G, Baudis M, Becker R, et al. Leveraging European infrastructures to access 1 million human genomes by 2022. Nat Rev Genet. 2019 Nov;20(11):693–701.