Toward the integration of Omics data in epidemiological studies: still a "long and winding road"

Publicly available datasets of breast histopathology H&E whole-slide images: A scoping review

Advancements in digital pathology and computing resources have made a significant impact in the field of computational pathology for breast cancer diagnosis and treatment. However, access to high-quality labeled histopathological images of breast cancer is a big challenge that limits the development of accurate and robust deep learning models. In this scoping review, we identified the publicly available datasets of breast H&E-stained whole-slide images (WSIs) that can be used to develop deep learning algorithms. We systematically searched 9 scientific literature databases and 9 research data repositories and found 17 publicly available datasets containing 10 385 H&E WSIs of breast cancer. Moreover, we reported image metadata and characteristics for each dataset to assist researchers in selecting proper datasets for specific tasks in breast cancer computational pathology. In addition, we compiled 2 lists of breast H&E patches and private datasets as supplementary resources for researchers. Notably, only 28% of the included articles utilized multiple datasets, and only 14% used an external validation set, suggesting that the performance of other developed models may be susceptible to overestimation. The TCGA-BRCA was used in 52% of the selected studies. This dataset has a considerable selection bias that can impact the robustness and generalizability of the trained algorithms. There is also a lack of consistent metadata reporting of breast WSI datasets that can be an issue in developing accurate deep learning models, indicating the necessity of establishing explicit guidelines for documenting breast WSI dataset characteristics and metadata.

The challenges of defining hormesis in epidemiological studies: The case of radiation hormesis

In the current radiation protection system, preventive measures and occupational exposure limits for controlling occupational exposure to ionizing radiation are based on the linear no-threshold extrapolation model. However, currently an increasing body of evidence indicates that this paradigm predicts very poorly biological responses in the low-dose exposure region. In addition, several in vitro and in vivo studies demonstrated the presence of hormetic dose response curves correlated to ionizing radiation low exposure. In this regard, it is noteworthy that also the findings of different epidemiological studies, conducted in different categories of occupationally exposed workers (e.g., healthcare, nuclear industrial and aircrew workers), observed lower rates of mortality and/or morbidity from cancer and/or other diseases in exposed workers than in unexposed ones or in the general population, then suggesting the possible occurrence of hormesis. Nevertheless, these results should be considered with caution since the identification of hormetic response in epidemiological studies is rather challenging because of a number of major limitations. In this regard, some of the most remarkable shortcomings found in epidemiological studies performed in workers exposed to ionizing radiation are represented by lack or inadequate definition of exposure doses, use of surrogates of exposure, narrow dose ranges, lack of proper control groups and poor evaluation of confounding factors. Therefore, considering the valuable role and contribution that epidemiological studies might provide to the complex risk assessment and management process, there is a clear and urgent need to overcome the aforementioned limits in order to achieve an adequate, useful and more real-life risk assessment that should also include the key concept of hormesis. Thus, in the present conceptual article we also discuss and provide possible approaches to improve the capacity of epidemiological studies to identify/define the hormetic response and consequently improve the complex process of risk assessment of ionizing radiation at low exposure doses.

Intracranial hemorrhage management in the multi-omics era

Intracranial hemorrhage (ICH) is a devastating disorder. Neuroprotective strategies that prevent tissue injury and improve functional outcomes have been identified in multiple animal models of ICH. However, these potential interventions in clinical trials produced generally disappointing results. With progress in omics, studies of omics data, including genomics, transcriptomics, epigenetics, proteomics, metabolomics, and the gut microbiome, may help promote precision medicine. In this review, we focused on introducing the applications of all omics in ICH and shed light on all of the considerable advantages to systematically analyze the necessity and importance of multiple omics technology in ICH.

Methods for Stratification and Validation Cohorts: A Scoping Review

Personalized medicine requires large cohorts for patient stratification and validation of patient clustering. However, standards and harmonized practices on the methods and tools to be used for the design and management of cohorts in personalized medicine remain to be defined. This study aims to describe the current state-of-the-art in this area. A scoping review was conducted searching in PubMed, EMBASE, Web of Science, Psycinfo and Cochrane Library for reviews about tools and methods related to cohorts used in personalized medicine. The search focused on cancer, stroke and Alzheimer's disease and was limited to reports in English, French, German, Italian and Spanish published from 2005 to April 2020. The screening process was reported through a PRISMA flowchart. Fifty reviews were included, mostly including information about how data were generated (25/50) and about tools used for data management and analysis (24/50). No direct information was found about the quality of data and the requirements to monitor associated clinical data. A scarcity of information and standards was found in specific areas such as sample size calculation. With this information, comprehensive guidelines could be developed in the future to improve the reproducibility and robustness in the design and management of cohorts in personalized medicine studies.

Network Analysis of Demographics, Dietary Intake, and Comorbidity Interactions

The aim of this study was to elucidate the complex interrelationships among dietary intake, demographics, and the risk of comorbidities. We applied a Gaussian graphical model to calculate the dietary scores of the participants. The network structure of dietary intake, demographics, and comorbidities was estimated in a mixed graphical model. The centrality indices of the nodes (strength (S), closeness (C), and betweenness (B)) were measured to identify the central node. Multinomial logistic regression was used to examine the association between the factors and comorbidities. Among 7423 participants, the strongest pairwise interactions were found between sex and smoking (1.56), sex and employment (0.66), sex and marital status (0.58), marital status and income (0.65), and age and employment (0.58). Among the factors in the network, sex played a central role (S = 4.63, C = 0.014, B = 41), followed by age (S = 2.81, C = 0.013, B = 18), smoking (S = 2.72, C = 0.013, B = 0), and employment (S = 2.17, C = 0.014, B = 22). While the odds of hypertension and diabetes were significantly higher among females than males, an inverse association was observed between high cholesterol and moderate chronic kidney disease. Among these factors, dietary intake was not a strongly interacting factor in the network, whereas age was consistently associated with the comorbidities of hypertension, high cholesterol, diabetes, and chronic kidney disease.

Network and Systems Medicine: Position Paper of the European Collaboration on Science and Technology Action on Open Multiscale Systems Medicine

Introduction: Network and systems medicine has rapidly evolved over the past decade, thanks to computational and integrative tools, which stem in part from systems biology. However, major challenges and hurdles are still present regarding validation and translation into clinical application and decision making for precision medicine. Methods: In this context, the Collaboration on Science and Technology Action on Open Multiscale Systems Medicine (OpenMultiMed) reviewed the available advanced technologies for multidimensional data generation and integration in an open-science approach as well as key clinical applications of network and systems medicine and the main issues and opportunities for the future. Results: The development of multi-omic approaches as well as new digital tools provides a unique opportunity to explore complex biological systems and networks at different scales. Moreover, the application of findable, applicable, interoperable, and reusable principles and the adoption of standards increases data availability and sharing for multiscale integration and interpretation. These innovations have led to the first clinical applications of network and systems medicine, particularly in the field of personalized therapy and drug dosing. Enlarging network and systems medicine application would now imply to increase patient engagement and health care providers as well as to educate the novel generations of medical doctors and biomedical researchers to shift the current organ- and symptom-based medical concepts toward network- and systems-based ones for more precise diagnoses, interventions, and ideally prevention. Conclusion: In this dynamic setting, the health care system will also have to evolve, if not revolutionize, in terms of organization and management.

Integrating biology and access to care in addressing breast cancer disparities: 25 years' research experience in the Carolina Breast Cancer Study

Purpose of Review

To review research on breast cancer mortality disparities, emphasizing research conducted in the Carolina Breast Cancer Study, with a focus on challenges and opportunities for integration of tumor biology and access characteristics across the cancer care continuum.

Recent Findings

Black women experience higher mortality following breast cancer diagnosis, despite lower incidence compared to white women. Biological factors, such as stage at diagnosis and breast cancer subtypes, play a role in these disparities. Simultaneously, social, behavioral, environmental, and access to care factors are important. However, integrated studies of biology and access are challenging and it is uncommon to have both data types available in the same study population. The central emphasis of Phase 3 of the Carolina Breast Cancer Study, initiated in 2008, was to collect rich data on biology (including germline and tumor genomics and pathology) and health care access in a diverse study population, with the long term goal of defining intervention opportunities to reduce disparities across the cancer care continuum.

Summary

Early and ongoing research from CBCS has identified important interactions between biology and access, leading to opportunities to build greater equity. However, sample size, population-specific relationships among variables, and complexities of treatment paths along the care continuum pose important research challenges. Interdisciplinary teams, including experts in novel data integration and causal inference, are needed to address gaps in our understanding of breast cancer disparities.

Network Aggregation to Enhance Results Derived from Multiple Analytics

The more complex data are, the higher the number of possibilities to extract partial information from those data. These possibilities arise by adopting different analytic approaches. The heterogeneity among these approaches and in particular the heterogeneity in results they produce are challenging for follow-up studies, including replication, validation and translational studies. Furthermore, they complicate the interpretation of findings with wide-spread relevance. Here, we take the example of statistical epistasis networks derived from genome-wide association studies with single nucleotide polymorphisms as nodes. Even though we are only dealing with a single data type, the epistasis detection problem suffers from many pitfalls, such as the wide variety of analytic tools to detect them, each highlighting different aspects of epistasis and exhibiting different properties in maintaining false positive control. To reconcile different network views to the same problem, we considered 3 network aggregation methods and discussed their performance in the context of epistasis network aggregation. We furthermore applied a latent class method as best performer to real-life data on inflammatory bowel disease (IBD) and highlighted its benefits to increase our understanding about IBD underlying genetic architectures.

Integrated analysis of genomics, longitudinal metabolomics, and Alzheimer's risk factors among 1,111 cohort participants

Although Alzheimer's disease (AD) is highly heritable, genetic variants are known to be associated with AD only explain a small proportion of its heritability. Genetic factors may only convey disease risk in individuals with certain environmental exposures, suggesting that a multiomics approach could reveal underlying mechanisms contributing to complex traits, such as AD. We developed an integrated network to investigate relationships between metabolomics, genomics, and AD risk factors using Wisconsin Registry for Alzheimer's Prevention participants. Analyses included 1,111 non-Hispanic Caucasian participants with whole blood expression for 11,376 genes (imputed from dense genome-wide genotyping), 1,097 fasting plasma metabolites, and 17 AD risk factors. A subset of 155 individuals also had 364 fastings cerebral spinal fluid (CSF) metabolites. After adjusting each of these 12,854 variables for potential confounders, we developed an undirected graphical network, representing all significant pairwise correlations upon adjusting for multiple testing. There were many instances of genes being indirectly linked to AD risk factors through metabolites, suggesting that genes may influence AD risk through particular metabolites. Follow-up analyses suggested that glycine mediates the relationship between carbamoyl-phosphate synthase 1 and measures of cardiovascular and diabetes risk, including body mass index, waist-hip ratio, inflammation, and insulin resistance. Further, 38 CSF metabolites explained more than 60% of the variance of CSF levels of tau, a detrimental protein that accumulates in the brain of AD patients and is necessary for its diagnosis. These results further our understanding of underlying mechanisms contributing to AD risk while demonstrating the utility of generating and integrating multiple omics data types.

Challenges in the Integration of Omics and Non-Omics Data

Omics data integration is already a reality. However, few omics-based algorithms show enough predictive ability to be implemented into clinics or public health domains. Clinical/epidemiological data tend to explain most of the variation of health-related traits, and its joint modeling with omics data is crucial to increase the algorithm’s predictive ability. Only a small number of published studies performed a “real” integration of omics and non-omics (OnO) data, mainly to predict cancer outcomes. Challenges in OnO data integration regard the nature and heterogeneity of non-omics data, the possibility of integrating large-scale non-omics data with high-throughput omics data, the relationship between OnO data (i.e., ascertainment bias), the presence of interactions, the fairness of the models, and the presence of subphenotypes. These challenges demand the development and application of new analysis strategies to integrate OnO data. In this contribution we discuss different attempts of OnO data integration in clinical and epidemiological studies. Most of the reviewed papers considered only one type of omics data set, mainly RNA expression data. All selected papers incorporated non-omics data in a low-dimensionality fashion. The integrative strategies used in the identified papers adopted three modeling methods: Independent, conditional, and joint modeling. This review presents, discusses, and proposes integrative analytical strategies towards OnO data integration.

Presidential address: Six open questions to genetic epidemiologists

Given the rapid pace with which genomics and other ‐omics disciplines are evolving, it is sometimes necessary to shift down a gear to consider more general scientific questions. In this line, in my presidential address I formulate six questions for genetic epidemiologists to ponder on. These cover the areas of reproducibility, statistical significance, chance findings, precision medicine and related fields such as bioinformatics and data science. Possible hints at responses are presented to foster our further discussion of these topics.

KUNO-Kids birth cohort study: rationale, design, and cohort description

Background

Birth cohort studies can contribute substantially to the understanding of health and disease — in childhood and over the life course. The KUNO-Kids birth cohort study was established to investigate various aspects of child health, using novel omics technologies in a systems medicine approach.

Results

After 3 years of recruitment, 2515 infants and their families have joined the study. Parents with higher education are overrepresented as in many other birth cohorts and are more likely to complete follow-up assessments via self-report questionnaires. The vast majority of participants consented to clinical examinations of their child and to the non-invasive collection of diverse biosamples, which were processed specifically for their integrated use in omics technology covering genomics, epigenomics, transcriptomics, metabolomics, and microbiome analyses of the skin, oral cavity, and stool.

Conclusions

The data and diverse biomaterial collected in the KUNO-Kids birth cohort study will provide extensive opportunities for investigating child health and its determinants in a holistic approach. The combination of a broad range of research questions in one study will allow for a cost-effective use of biomaterial and omics results and for a comprehensive analysis of biological and social determinants of health and disease. Aiming for low attrition and ensuring participants’ long-term commitment will be crucial to fully exploit the potential of the study.

Causal Inference in Cancer Epidemiology: What Is the Role of Mendelian Randomization?

Observational epidemiologic studies are prone to confounding, measurement error, and reverse causation, undermining robust causal inference. Mendelian randomization (MR) uses genetic variants to proxy modifiable exposures to generate more reliable estimates of the causal effects of these exposures on diseases and their outcomes. MR has seen widespread adoption within cardio-metabolic epidemiology, but also holds much promise for identifying possible interventions for cancer prevention and treatment. However, some methodologic challenges in the implementation of MR are particularly pertinent when applying this method to cancer etiology and prognosis, including reverse causation arising from disease latency and selection bias in studies of cancer progression. These issues must be carefully considered to ensure appropriate design, analysis, and interpretation of such studies. In this review, we provide an overview of the key principles and assumptions of MR, focusing on applications of this method to the study of cancer etiology and prognosis. We summarize recent studies in the cancer literature that have adopted a MR framework to highlight strengths of this approach compared with conventional epidemiological studies. Finally, limitations of MR and recent methodologic developments to address them are discussed, along with the translational opportunities they present to inform public health and clinical interventions in cancer. Cancer Epidemiol Biomarkers Prev; 27(9); 995-1010. ©2018 AACR.

Variable selection in omics data: A practical evaluation of small sample sizes

In omics experiments, variable selection involves a large number of metabolites/ genes and a small number of samples (the n < p problem). The ultimate goal is often the identification of one, or a few features that are different among conditions- a biomarker. Complicating biomarker identification, the p variables often contain a correlation structure due to the biology of the experiment making identifying causal compounds from correlated compounds difficult. Additionally, there may be elements in the experimental design (blocks, batches) that introduce structure in the data. While this problem has been discussed in the literature and various strategies proposed, the over fitting problems concomitant with such approaches are rarely acknowledged. Instead of viewing a single omics experiment as a definitive test for a biomarker, an unrealistic analytical goal, we propose to view such studies as screening studies where the goal of the study is to reduce the number of features present in the second round of testing, and to limit the Type II error. Using this perspective, the performance of LASSO, ridge regression and Elastic Net was compared with the performance of an ANOVA via a simulation study and two real data comparisons. Interestingly, a dramatic increase in the number of features had no effect on Type I error for the ANOVA approach. ANOVA, even without multiple test correction, has a low false positive rates in the scenarios tested. The Elastic Net has an inflated Type I error (from 10 to 50%) for small numbers of features which increases with sample size. The Type II error rate for the ANOVA is comparable or lower than that for the Elastic Net leading us to conclude that an ANOVA is an effective analytical tool for the initial screening of features in omics experiments.

pulver: an R package for parallel ultra-rapid p-value computation for linear regression interaction terms

Background

Genome-wide association studies allow us to understand the genetics of complex diseases. Human metabolism provides information about the disease-causing mechanisms, so it is usual to investigate the associations between genetic variants and metabolite levels. However, only considering genetic variants and their effects on one trait ignores the possible interplay between different “omics” layers. Existing tools only consider single-nucleotide polymorphism (SNP)–SNP interactions, and no practical tool is available for large-scale investigations of the interactions between pairs of arbitrary quantitative variables.

Results

We developed an R package called pulver to compute p-values for the interaction term in a very large number of linear regression models. Comparisons based on simulated data showed that pulver is much faster than the existing tools. This is achieved by using the correlation coefficient to test the null-hypothesis, which avoids the costly computation of inversions. Additional tricks are a rearrangement of the order, when iterating through the different “omics” layers, and implementing this algorithm in the fast programming language C++. Furthermore, we applied our algorithm to data from the German KORA study to investigate a real-world problem involving the interplay among DNA methylation, genetic variants, and metabolite levels.

Conclusions

The pulver package is a convenient and rapid tool for screening huge numbers of linear regression models for significant interaction terms in arbitrary pairs of quantitative variables. pulver is written in R and C++, and can be downloaded freely from CRAN at https://cran.r-project.org/web/packages/pulver/.

Electronic supplementary material

The online version of this article (10.1186/s12859-017-1838-y) contains supplementary material, which is available to authorized users.

Trends in the application of high-resolution mass spectrometry for human biomonitoring: An analytical primer to studying the environmental chemical space of the human exposome

Global profiling of xenobiotics in human matrices in an untargeted mode is gaining attention for studying the environmental chemical space of the human exposome. Defined as the study of a comprehensive inclusion of environmental influences and associated biological responses, human exposome science is currently evolving out of the metabolomics science. In analogy to the latter, the development and applications of high resolution mass spectrometry (HRMS) has shown potential and promise to greatly expand our ability to capture the broad spectrum of environmental chemicals in exposome studies. HRMS can perform both untargeted and targeted analysis because of its capability of full- and/or tandem-mass spectrum acquisition at high mass accuracy with good sensitivity. The collected data from target, suspect and non-target screening can be used not only for the identification of environmental chemical contaminants in human matrices prospectively but also retrospectively. This review covers recent trends and advances in this field. We focus on advances and applications of HRMS in human biomonitoring studies, and data acquisition and mining. The acquired insights provide stepping stones to improve understanding of the human exposome by applying HRMS, and the challenges and prospects for future research.