Next Chapter of the Taiwan Biobank: Sustainability and Perspectives

Comparative analysis of machine learning algorithms for Alzheimer's disease classification using EEG signals and genetic information

Alzheimer's disease (AD) is a progressive neurodegenerative disorder characterized by cognitive decline, memory impairments, and behavioral changes. The presence of abnormal beta-amyloid plaques and tau protein tangles in the brain is known to be associated with AD. However, current limitations of imaging technology hinder the direct detection of these substances. Consequently, researchers are exploring alternative approaches, such as indirect assessments involving monitoring brain signals, cognitive decline levels, and blood biomarkers. Recent studies have highlighted the potential of integrating genetic information into these approaches to enhance early detection and diagnosis, offering a more comprehensive understanding of AD pathology beyond the constraints of existing imaging methods. Our study utilized electroencephalography (EEG) signals, genotypes, and polygenic risk scores (PRSs) as features for machine learning models. We compared the performance of gradient boosting (XGB), random forest (RF), and support vector machine (SVM) to determine the optimal model. Statistical analysis revealed significant correlations between EEG signals and clinical manifestations, demonstrating the ability to distinguish the complexity of AD from other diseases by using genetic information. By integrating EEG with genetic data in an SVM model, we achieved exceptional classification performance, with an accuracy of 0.920 and an area under the curve of 0.916. This study presents a novel approach of utilizing real-time EEG data and genetic background information for multimodal machine learning. The experimental results validate the effectiveness of this concept, providing deeper insights into the actual condition of patients with AD and overcoming the limitations associated with single-oriented data.

Sex Disparity in the Association of Metabolic Syndrome with Cognitive Impairment

Background/Objectives: Metabolic syndrome (MS) is a constellation of several cardiometabolic risk factors. We investigated sex disparity in the associations between MS and cognitive impairment using cross-sectional data from Taiwan Biobank. Methods: We determined the associations of MS and its five components with cognitive impairment (mini-mental state examination, MMSE < 24) and the five domains of MMSE using logistic regression analyses. Results: A total of 7399 men and 11,546 women were included, and MS was significantly associated with cognitive impairment only in women (adjusted OR 1.48, 95% CI 1.29-1.71, p = 0.001) (p for interaction 0.005). In women, the association with MS was significant in orientation (adjusted OR 1.21, 95% CI 1.07-1.37, p = 0.003), memory (adjusted OR 1.12, 95% CI 1.01-1.25, p = 0.034) and design copying (adjusted OR 1.41, 95% CI 1.23-1.62, p = 0.001) (p value for interaction 0.039, 0.023, and 0.093, respectively). Among the components of MS, a large waist circumference (adjusted OR 1.25, 95% CI 1.08-1.46, p = 0.003), high fasting glucose (adjusted OR 1.16, 95% CI 1.00-1.34, p = 0.046), and low HDL cholesterol (adjusted OR 1.16, 95% CI 1.00-1.34, p = 0.049) were significantly associated with cognitive impairment in women. Conclusions: Our findings suggest that sex has a significant influence on the association between MS and cognitive dysfunction, especially in orientation and memory.

A Rapid Review on the Value of Biobanks Containing Genetic Information

OBJECTIVES

Increasing access to health data through biobanks containing genetic information has the potential to expand the knowledge base and thereby improve screening, diagnosis, and treatment options for many diseases. Nevertheless, although privacy concerns and risks surrounding genetic data sharing are well documented, direct evidence in favor of the hypothesized benefits of data integration is scarce, which complicates decision making in this area. Therefore, the objective of this study is to summarize the available evidence on the research and clinical impacts of biobanks containing genetic information, so as to better understand how to quantify the value of expanding genomic data access.

METHODS

Using a rapid review methodology, we performed a search of MEDLINE/PubMed and Embase databases; and websites of biobanks and genomic initiatives published from 2010 to 2022. We classified findings into 11 indicators including outputs (a direct product of the biobank activities) and outcomes (changes in scientific and clinical capacity).

RESULTS

Of 8479 abstracts and 101 gray literature sources were reviewed, 96 records were included. Although most records did not report key indicators systematically, the available evidence concentrated on research indicators such as publications and gene-disorder association discoveries (63% of studies), followed by research infrastructure (26%), and clinical indicators (11%) such as supporting the diagnosis of individual patients.

CONCLUSIONS

Existing evidence on the benefits of biobanks is skewed toward easily quantifiable research outputs. Measuring a comprehensive set of outputs and outcomes inspired by value frameworks is necessary to generate better evidence on the benefits of genomic data sharing.

Development and validation of an insulin resistance model for a population without diabetes mellitus and its clinical implication: a prospective cohort study

Background

Insulin resistance (IR) is associated with diabetes mellitus, cardiovascular disease (CV), and mortality. Few studies have used machine learning to predict IR in the non-diabetic population.

Methods

In this prospective cohort study, we trained a predictive model for IR in the non-diabetic populations using the US National Health and Nutrition Examination Survey (NHANES, from JAN 01, 1999 to DEC 31, 2012) database and the Taiwan MAJOR (from JAN 01, 2008 to DEC 31, 2017) database. We analysed participants in the NHANES and MAJOR and participants were excluded if they were aged <18 years old, had incomplete laboratory data, or had DM. To investigate the clinical implications (CV and all-cause mortality) of this trained model, we tested it with the Taiwan biobank (TWB) database from DEC 10, 2008 to NOV 30, 2018. We then used SHapley Additive exPlanation (SHAP) values to explain differences across the machine learning models.

Findings

Of all participants (combined NHANES and MJ databases), we randomly selected 14,705 participants for the training group, and 4018 participants for the validation group. In the validation group, their areas under the curve (AUC) were all >0.8 (highest being XGboost, 0.87). In the test group, all AUC were also >0.80 (highest being XGboost, 0.88). Among all 9 features (age, gender, race, body mass index, fasting plasma glucose (FPG), glycohemoglobin, triglyceride, total cholesterol and high-density cholesterol), BMI had the highest value of feature importance on IR (0.43 for XGboost and 0.47 for RF algorithms). All participants from the TWB database were separated into the IR group and the non-IR group according to the XGboost algorithm. The Kaplan-Meier survival curve showed a significant difference between the IR and non-IR groups (p < 0.0001 for CV mortality, and p = 0.0006 for all-cause mortality). Therefore, the XGboost model has clear clinical implications for predicting IR, aside from CV and all-cause mortality.

Interpretation

To predict IR in non-diabetic patients with high accuracy, only 9 easily obtained features are needed for prediction accuracy using our machine learning model. Similarly, the model predicts IR patients with significantly higher CV and all-cause mortality. The model can be applied to both Asian and Caucasian populations in clinical practice.

Funding

Taichung Veterans General Hospital, Taiwan and Japan Society for the Promotion of Science KAKENHI Grant Number JP21KK0293.

Challenges on the Development of Biodiversity Biobanks: The Living Archives of Biodiversity

Biodiversity biobanks or ex situ biodiversity biorepositories tend to receive less attention compared with their biomedical counterparts. In this review, we highlight the necessity for these biorepositories by presenting their significant role in health, biodiversity, linking of big data, other translational research, and biodiversity conservation efforts. Moreover, the significant challenges in developing and maintaining biodiversity biobanks based on successful biobanks in some megadiverse developing countries are examined to provide insights into what needs to be done and what can be improved by up-and-coming biodiversity biobanks. These challenges include lack of financial support and political will; availability of experts; development of standard policies; and information management system. In addition, issues regarding access and benefit sharing and Digital Sequence Information must be addressed by biodiversity biobanks.

Economics of Biobanking: Business or Public Good? Literature Review, Structural and Thematic Analysis

This paper reviewed the relevant scientific literature on the business and economics of biobanking to explore key themes and paradigms. The structural properties of the literature were investigated, such as key authors, journals, studies, as well as co-citation and co-authorship networks; the study revealed that the research on business and economics is a niche area within the vast biobanking literature. The research is concentrated in a relatively small number of journals, institutions, and countries, which is rather surprising given the substantial public investment in and concerns about biobank sustainability. The structural analysis also suggested major themes in research on biobanking business and economics and noted shifts in focus on specific themes. The commercialisation of samples is more acknowledged than before but under the condition of equitable sharing of benefits across various stakeholders. Most biobanks are heavily subsidised by the public sector and are considered public goods rather than business enterprises. This is OK, but underutilisation of specimens and low rates of cost recovery suggest that the current mainstream operating model is hardly sustainable. With many biobanks maturing, long-term sustainability became a key topic of the discussion on biobanking trends.

Precision Health in Taiwan: A Data-Driven Diagnostic Platform for the Future of Disease Prevention

“Precision medicine” has revolutionized how we respond to diseases by using an individual’s genomic data and lifestyle and environment-related information to create an effective personalized treatment. However, issues surrounding regulations, medical insurance payments and the use of patients’ medical data, have delayed the development of precision medicine and made it difficult to achieve “true” personalization. We therefore recommend that precision medicine be transformed into precision health: a novel and generalized platform of tools and methods that could prevent, manage, and treat disease at a population level. “Precision health,” one of six core strategic industries highlighted in Taiwan’s vision for 2030, uses various physiological data, genomic data, and external factors, to develop unique “preventative” solutions or therapeutic strategies. For Taiwan to implement precision health, it has to address three challenges: (1) the high-cost issue of precision health; (2) the harmonization issues surrounding integration and transmission of specimen and data; (3) the legal issue of combining information and communications technology (ICT) with Artificial Intelligence (AI) for medical use. In this paper, we propose an innovative framework with six recommendations for facilitating the development of precision health in Taiwan, including a novel model of precise telemedicine with AI-aided technology. We then describe how these tools can be proactively applied in early response to the COVID-19 crisis. We believe that precision health represents an important shift to more proactive and preventive healthcare that enables people to lead healthier lives.

The Association between Bodily Pain and Cognitive Impairment in Community-Dwelling Older Adults

Background: Bodily pain is a common condition in older adults and interferes with individuals’ cognitive functioning. We aimed to evaluate the association between bodily pain and related locations and cognitive impairment among community-dwelling older adults in Taiwan. Method: In this retrospective, cross-sectional study, we enrolled 2022 participants aged 60‒70 years, from the Taiwan Biobank. Mini-Mental State Examination was performed to assess cognitive impairment. Further, logistic regression analyses were performed to identify the relationship between bodily pain and cognitive impairment. Results: Overall, 161 participants had cognitive impairment. Multivariable analysis showed that older adults who reported bodily pain were more likely than those who did not have cognitive impairment (odds ratio 1.68). Moreover, the occurrence of cognitive impairment correlated with the presence of two or more pain locations and self-reported low back and waist pain or sciatica. Conclusion: Our study revealed that cognitive impairment was associated with bodily pain in community-dwelling older adults, particularly older adults with low back and waist pain or sciatica and those with two or more pain locations. To maintain the quality of older adults’ life, pain and cognitive decline need to be simultaneously assessed with considerably more precise and objective markers.

Artificial-Intelligence-Assisted Discovery of Genetic Factors for Precision Medicine of Antiplatelet Therapy in Diabetic Peripheral Artery Disease

An increased risk of cardiovascular events was identified in patients with peripheral artery disease (PAD). Clopidogrel is one of the most widely used antiplatelet medications. However, there are heterogeneous outcomes when clopidogrel is used to prevent cardiovascular events in PAD patients. Here, we use an artificial intelligence (AI)-assisted methodology to identify genetic factors potentially involved in the clopidogrel-resistant mechanism, which is currently unclear. Several discoveries can be pinpointed. Firstly, a high proportion (>50%) of clopidogrel resistance was found among diabetic PAD patients in Taiwan. Interestingly, our result suggests that platelet function test-guided antiplatelet therapy appears to reduce the post-interventional occurrence of major adverse cerebrovascular and cardiac events in diabetic PAD patients. Secondly, AI-assisted genome-wide association study of a single-nucleotide polymorphism (SNP) database identified a SNP signature composed of 20 SNPs, which are mapped into 9 protein-coding genes (SLC37A2, IQSEC1, WASHC3, PSD3, BTBD7, GLIS3, PRDM11, LRBA1, and CNR1). Finally, analysis of the protein connectivity map revealed that LRBA, GLIS3, BTBD7, IQSEC1, and PSD3 appear to form a protein interaction network. Intriguingly, the genetic factors seem to pinpoint a pathway related to endocytosis and recycling of P2Y12 receptor, which is the drug target of clopidogrel. Our findings reveal that a combination of AI-assisted discovery of SNP signatures and clinical parameters has the potential to develop an ethnic-specific precision medicine for antiplatelet therapy in diabetic PAD patients.

Interaction of Alcohol Consumption and ABCG2 rs2231142 Variant Contributes to Hyperuricemia in a Taiwanese Population

Background: ABCG2 rs2231142 is an important genetic factor that contributes to the development of gout and hyperuricemia (HUA). Epidemiologic studies have demonstrated that lifestyle risk factors of HUA (e.g., alcohol consumption) and genetic predisposition (e.g., ABCG2 gene) together, contribute to enhanced serum uric acid levels. However, the interaction between ABCG2 rs2231142, alcohol consumption, and HUA in the Taiwanese population is still unclear. Therefore, this study investigated whether the risk of HUA is associated with ABCG2 rs2231142 variants and how this is affected by alcohol consumption. Method: study subjects were selected from the participants of the Taiwan Biobank database. Overall, 114,540 participants aged 30 to 70 years were enrolled in this study. The interaction between ABCG2 rs2231142, alcohol consumption, and serum uric acid (sUA) levels was analyzed by multiple logistic regression models. Results: the prevalence of HUA was 32.7% and 4.4 % in the male and female populations, respectively. In the whole study population, the minor T allele of ABCG2 rs2231142 was significantly associated with HUA risk, and the occurrence of HUA was high in TT genotype and TG genotype. The risk of HUA was significantly increased by the combined association of ABCG2 rs2231142 and alcohol consumption for TG/TT genotype compared to the GG genotype (wild-type genotype), especially among women. Conclusion: the ABCG2 rs2231142 is a crucial genetic locus for sUA levels in the Taiwanese population and our findings revealed that alcohol consumption combined with the ABCG2 rs2231142 risk allele contributes to increased HUA risk.

A Novel Quality-Control Procedure to Improve the Accuracy of Rare Variant Calling in SNP Arrays

Background: Single-nucleotide polymorphism (SNP) arrays are an ideal technology for genotyping genetic variants in mass screening. However, using SNP arrays to detect rare variants [with a minor allele frequency (MAF) of <1%] is still a challenge because of noise signals and batch effects. An approach that improves the genotyping quality is needed for clinical applications. Methods: We developed a quality-control procedure for rare variants which integrates different algorithms, filters, and experiments to increase the accuracy of variant calling. Using data from the TWB 2.0 custom Axiom array, we adopted an advanced normalization adjustment to prevent false calls caused by splitting the cluster and a rare het adjustment which decreases false calls in rare variants. The concordance of allelic frequencies from array data was compared to those from sequencing datasets of Taiwanese. Finally, genotyping results were used to detect familial hypercholesterolemia (FH), thrombophilia (TH), and maturity-onset diabetes of the young (MODY) to assess the performance in disease screening. All heterozygous calls were verified by Sanger sequencing or qPCR. The positive predictive value (PPV) of each step was estimated to evaluate the performance of our procedure. Results: We analyzed SNP array data from 43,433 individuals, which interrogated 267,247 rare variants. The advanced normalization and rare het adjustment methods adjusted genotyping calling of 168,134 variants (96.49%). We further removed 3916 probesets which were discordant in MAFs between the SNP array and sequencing data. The PPV for detecting pathogenic variants with 0.01%10,000 are available. The results demonstrated our procedure could perform correct genotype calling of rare variants. It provides a solution of pathogenic variant detection through SNP array. The approach brings tremendous promise for implementing precision medicine in medical practice.

Association of ABCG2 rs2231142 Allele and BMI With Hyperuricemia in an East Asian Population

Objectives: Genetic variants and obesity are risk factors for hyperuricemia (HUA). Recent genome-wide association studies have identified ABCG2 rs2231142 as one of the most prominent genetic variants for HUA in an East Asian population. Nevertheless, no large-scale studies have demonstrated any interactive effects between this variant and obesity on serum urate level in Asians. This study aimed to determine the interaction of ABCG2 rs2231142 variant and body mass index (BMI) and its effect on risk of HUA in an East Asian population.

Methods: The study was conducted using the Taiwan Biobank database, a population-based biomedical research database of patients with Taiwanese Han Chinese ancestry aged 30–70years between September 2014 and May 2017. Detailed physical information on participants were collected by questionnaires and genotyping using Affymetrix TWB 650K SNP chip. The primary outcome was HUA, defined as a serum uric acid level>7.0mg/dl. Odds ratio (OR) of HUA was analyzed using logistic regression models and the effects of interaction between ABCG2 rs2231142 variants and BMI on serum uric acid level were explored.

Results: We identified 25,245 subjects, 4,228 (16.75%) of whom had HUA. The prevalence of HUA was 30% in men and 3.8% in women. The risk of HUA was significantly associated with ABCG2 rs2231142 risk T allele, with more HUA in TT genotype (OR: 2.40, 95% CI: 2.11–2.72, p<0.001) and TG genotype (OR: 1.64, 95% CI: 1.51–1.78, p<0.001) in men, and TT genotype (OR: 2.42, 95% CI: 1.83–3.20, p<0.001) and TG genotype (OR: 1.82, 95% CI: 1.46–2.23, p<0.001) in women, compared with their counterparts. Moreover, we found a strong genetic-environmental interaction associated with the risk of HUA. There was increased risk of HUA by the interaction of ABCG2 rs2231142 variant and BMI for TT genotype (OR: 7.42, 95% CI: 2.54–21.7, p<0.001) and TG genotype (OR: 4.25, 95% CI: 2.13–8.47, p<0.001) in men compared with the GG genotype in men, and for TT genotype (OR: 25.43, 95% CI: 3.75–172.41, p<0.001) and TG genotype (OR: 3.05, 95% CI: 0.79–11.71, p=0.011) in women compared with the GG genotype in women.

Conclusion: The risk of HUA was markedly increased by the interaction of ABCG2 rs2231142 variant and BMI, both in men and in women. Body weight control and reduction in BMI are recommended in high-risk patients with the ABCG2 rs2231142 risk T allele.

Phage Biobank: Present Challenges and Future Perspectives

After a century of use in human infection, the preparation and administration of therapeutic bacteriophages (phages) still relies on ad hoc partnerships of researchers, biotech companies, clinicians and regulators. There is a clear need to improve the reproducibility, safety and speed of the provision of suitable phages. Here we discuss the specific characteristics and challenges of a sustainable phage biobank and, as we build a national consortium aimed at delivering phage therapeutics, suggest a roadmap toward national biobanking and phage therapy initiatives using the Australian context as a model.

Practical guide for managing large-scale human genome data in research

Studies in human genetics deal with a plethora of human genome sequencing data that are generated from specimens as well as available on public domains. With the development of various bioinformatics applications, maintaining the productivity of research, managing human genome data, and analyzing downstream data is essential. This review aims to guide struggling researchers to process and analyze these large-scale genomic data to extract relevant information for improved downstream analyses. Here, we discuss worldwide human genome projects that could be integrated into any data for improved analysis. Obtaining human whole-genome sequencing data from both data stores and processes is costly; therefore, we focus on the development of data format and software that manipulate whole-genome sequencing. Once the sequencing is complete and its format and data processing tools are selected, a computational platform is required. For the platform, we describe a multi-cloud strategy that balances between cost, performance, and customizability. A good quality published research relies on data reproducibility to ensure quality results, reusability for applications to other datasets, as well as scalability for the future increase of datasets. To solve these, we describe several key technologies developed in computer science, including workflow engine. We also discuss the ethical guidelines inevitable for human genomic data analysis that differ from model organisms. Finally, the future ideal perspective of data processing and analysis is summarized.

Transformation of the Taiwan Biobank 3.0: vertical and horizontal integration

Researchers expect a high quality of biospecimens/data and value-added services from biobanks. Therefore, the concept of "biobank 3.0" was introduced so that biobanks could better meet the needs of stakeholders and maintain sustainable operations. Theoretically, the Taiwan Biobank (TWB) has already gone through the concepts of biobank 1.0 and 2.0. However, three challenges still need to be addressed before it can be transformed into a new generation of the TWB (namely, the TWB 3.0): (1) the difficulty of integrating other biobanks' resources, (2) the efficiency and effectiveness of the release and use of biospecimens/data, and (3) the development of income and revenue models of sustainability. To address these issues, this paper proposes a framework for the TWB 3.0 transformation based on a dual-pillar approach composed of a "physically" vertical integration driven by the TWB and a "virtually" horizontal network led by the National Health Research Institutes (NHRI) of Taiwan. Using prominent biobanks such as the Biobanking and BioMolecular Resources Research Infrastructure-European Research Infrastructure Consortium (BBMRI-ERIC), the UK Biobank, and the National Institutes of Health (NIH)'s All of Us Research Program as models, the TWB can strengthen its on-going TWB 2.0 operations in regional and/or international collaboration, increase the value of data collected and develop closer relationships with biobank participants and users. To these ends, the authors highlight key issues that include, but are not limited to, the harmonization of relevant ELSI standards for various biobanks' integrations; the value-added services and the efficiency of Big Data Era related research and/or precision medicine development, and financial concerns related to biobank sustainability. This paper concludes by discussing how greater participant engagement and the uptake of Information Technology (IT) and Artificial Intelligence (AI) applications can be used in partnership with vertical and horizontal integration as part of a four-pronged approach to promote biobank sustainability, and facilitate the TWB 3.0 transformation.

Mini-Review of Laboratory Operations in Biobanking: Building Biobanking Resources for Translational Research

Biobanks have become integral to improving population health. We are in a new era in medicine as patients, health professionals, and researchers increasingly collaborate to gain new knowledge and explore new paradigms for diagnosing and treating disease. Many large-scale biobanking efforts are underway worldwide at the institutional, national, and even international level. When linked with subject data from questionnaires and medical records, biobanks serve as valuable resources in translational research. A biobank must have high quality samples that meet researcher's needs. Biobank laboratory operations require an enormous amount of support—from lab and storage space, information technology expertise, and a laboratory management information system to logistics for sample movement, quality management systems, and appropriate facilities. A paramount metric of success for a biobank is the concept of every biospecimen coming to the repository belongs to a participant who has something to contribute to research for a healthier future. This article will discuss the importance of biorepository operations, specific to the collection and storage of participants materials. Specific focus will be given to maintaining the quality of samples, along with the various levels of support biorepositories need to fulfill their purpose and ensure the integrity of each specimen is maintained.

Managing "incidental findings" in biobank research: Recommendations of the Taiwan biobank

BACKGROUND

In this article, incidental findings (IF) refer to unforeseen findings made possible through biobanking research and advances in medical diagnostic technologies that raise issues regarding the obligation and/or responsibility of biobank-users and biobanks to return clinically significant information to participants. The World Medical Association (WMA) Declaration of Taipei (2016) highlights the possibility of encountering IF and requires that research on biospecimens address biobank feedback policies in their informed consent process, leaving open the possibility that the policy may be "no return". As clinicians and researchers begin to use these "resources", the possibility of finding clinically significant IF is becoming a reality.

DISCUSSION

In line with the WMA's Declaration of Taipei, a pragmatic approach is needed to deal with the issue of returning IF in biobank governance. Indeed, the impacts and concerns associated with the return of IF differ across different stakeholder groups and jurisdictions. Therefore, the framework governing IF return needs to be custom-built, taking into account the nature of each research project and the unique features of biobanks. To this end, in addition to facilitating biobank transparency, establishing an endurable and horizontal connection among biobanks and clinical institutions under a public health system will improve efficiency and effectiveness. Hence, subject to contemporary Taiwanese ethical and/or legal regulations, this article argues for the establishment of an updated framework for imaging-related and genetic-related IF return within the Taiwan Biobank (TWB), mainly based on a limited obligation to disclose life-threatening information revealed by imaging, but not genetic, information.

SUMMARY

After discussing some of the ethical, legal and social issues encountered by the TWB and accounting for the experiences of other international biobanks, we propose a systematic framework for returning IF, mainly on a "limited obligation" basis, which offers better and more comprehensive protection for biobank-participants' rights and health.