神经精神疾病中全基因组DNA甲基化研究的综合指南及其对深空环境的启示（英文）

A comprehensive guide to genome-wide DNA methylation research in neuropsychiatric disorders and its implications for deep-space environments

摘要

神经精神疾病源于遗传与环境因素之间的复杂相互作用。DNA甲基化作为一种可逆且对环境高度敏感的表观遗传调控机制，在环境暴露、基因表达调节及神经行为结果之间发挥关键的桥梁作用。在长期的深空航天任务中，宇航员同时暴露于微重力、宇宙辐射、昼夜节律紊乱及社交隔离等多重应激因素，易发生DNA甲基化及神经精神疾病。全基因组DNA甲基化研究可分为研究设计、样本处理与检测、数据分析3个主要方法学阶段，可应用于宇航员神经精神健康监测。Illumina MethylationEPIC芯片与全基因组亚硫酸氢盐测序2种技术在覆盖度、分辨率、成本及研究适用性等方面存在优势互补：芯片方法成本较低，适用于大样本人群研究和长期随访监测；测序方法虽成本较高，但具有更高的覆盖度与分辨率，更适合高精度甲基化图谱构建及个体差异分析；此外，单细胞甲基化测序、纳米孔长读长测序及基于机器学习的多组学整合等新兴技术的引入，有望进一步提升表观遗传研究的精准性与解释力。这些方法学的进展可为构建基于DNA甲基化的宇航员神经精神风险监测体系提供关键支撑，并为未来长期深空任务的神经精神健康保障奠定了表观遗传学基础。

Abstract

Neuropsychiatric disorders arise from complex interactions between genetic and environmental factors. DNA methylation, a reversible and environmentally responsive epigenetic regulatory mechanism, serves as a crucial bridge linking environmental exposure, gene expression regulation, and neurobehavioral outcomes. During long-duration deep-space missions, astronauts face multiple stressors-including microgravity, cosmic radiation, circadian rhythm disruption, and social isolation, which can induce alterations in DNA methylation and increase the risk of neuropsychiatric disorders. Genome-wide DNA methylation research can be divided into 3 major methodological stages: Study design, sample preparation and detection, and data analysis, each of which can be applied to astronaut neuropsychiatric health monitoring. Systematic comparison of the Illumina MethylationEPIC array and whole-genome bisulfite sequencing reveals their complementary strengths in terms of genomic coverage, resolution, cost, and application scenarios: the array method is cost-effective and suitable for large-scale population studies and longitudinal monitoring, whereas sequencing provides higher resolution and coverage and is more suitable for constructing detailed methylation maps and characterizing individual variation. Furthermore, emerging technologies such as single-cell methylation sequencing, nanopore long-read sequencing, and machine-learning-based multi-omics integration are expected to greatly enhance the precision and interpretability of epigenetic studies. These methodological advances provide key support for establishing DNA-methylation-based monitoring systems for neuropsychiatric risk in astronauts and lay an epigenetic foundation for safeguarding neuropsychiatric health during future long-term deep-space missions.

关键词

神经精神疾病 / DNA甲基化 / 表观遗传学 / 深空飞行 / MethylationEPIC芯片 / 全基因组亚硫酸氢盐测序

Key words

neuropsychiatric disorders / DNA methylation / epigenetics / deep-space flight / MethylationEPIC array / whole-genome bisulfite sequencing

As global space agencies accelerate plans for manned deep space missions, the potential impact of prolonged spaceflight on astronauts’ physical and mental health has become a pressing concern in aerospace medicine. Unique environmental stressors in space-such as microgravity, cosmic radiation, circadian rhythm disruption, and psychosocial isolation-have been shown to interfere with multiple physiological systems, with the central nervous system being particularly susceptible^[1]. Findings from animal studies^[2-3], ground-based analogs^[4-5], and in-orbit missions^[6] indicate that these extreme conditions may induce depressive-like behaviors, anxiety, and cognitive deficits, thereby increasing the risk of neuropsychiatric disorders (NPDs). Epigenetic regulation, particularly DNA methylation, has emerged as a key molecular interface linking environmental exposures to long-term neurobehavioral alterations^[7-12].

NPDs encompass a diverse spectrum of conditions, including schizophrenia (SCZ), bipolar disorder (BD), autism spectrum disorder (ASD), Alzheimer’s disease (AD), and Parkinson’s disease (PD). These disorders frequently lead to substantial impairments in cognition, emotion, and social functioning. Their etiology involves complex interactions between genetic predispositions and environmental influences^[13-15]. For example, monozygotic (MZ) twins-who share nearly identical genomes-often show discordant phenotypes; a Danish twin study^[16] reported only a 33% concordance rate for SCZ, underscoring the significance of non-genetic factors. DNA methylation regulates gene expression without altering the DNA sequence and serves as a core mechanism through which environmental stimuli exert their biological effects^{[10, 17-18]}. While unmethylated CpG islands in promoter regions usually facilitate transcription, gene body methylation often correlates positively with expression levels^[19-21]. Dysregulation of these patterns has been implicated in the pathophysiology of various NPDs, affecting neurodevelopment, synaptic plasticity, and neuroimmune homeostasis^[7-12]. Disorder-specific methylation profiles may function not only as diagnostic or prognostic biomarkers, but also as tools for monitoring disease progression, evaluating therapeutic responses, and identifying at-risk individuals^{[11, 22-27]}.

Recent advances^[2-6] in space biomedical research have revealed that space exposure can significantly alter the human methylome. Landmark studies such as the NASA Twins Study^[6], the Mars500 project^[5], and rodent spaceflight experiments^[2-3] have demonstrated that microgravity and space radiation can induce persistent methylation changes in genes associated with immune responses, oxidative stress, and neuroinflammation. These findings^[2-6] highlight the exceptional sensitivity of the epigenome to environmental perturbations and point to its potential utility in managing mental health risks during long-duration missions.

Despite growing interest, research on DNA methylation in NPDs still faces numerous conceptual and technical challenges, including cellular heterogeneity, platform selection, and data interpretation. Advances in high-throughput technologies, particularly the Illumina MethylationEPIC array and whole-genome bisulfite sequencing (WGBS), have enabled systematic exploration of the methylome under both terrestrial and spaceflight conditions. Against this backdrop, this review outlines key methodological considerations for genome-wide DNA methylation studies, including study design and participant recruitment, sample processing and detection, and downstream data analysis. It further compares the strengths and limitations of the MethylationEPIC array and WGBS platforms, and discusses recent innovations such as single-cell methylation profiling, machine learning-based risk models, the T2T-CHM13 reference genome, and nanopore sequencing. This review aims to provide an essential reference point for extending epigenetic research into spaceflight conditions and astronaut mental health monitoring.

1 Bisulfite conversion-based detection approaches

1.1 Bisulfite conversion

Genome-wide DNA methylation studies typically employ 1 of 3 experimental approaches, depending on the method of DNA treatment: Bisulfite (BS) conversion-based, restriction enzyme-based, and affinity enrichment-based methods^[28]. Among these, BS conversion is the most widely used. This approach involves treating genomic DNA with sodium BS, which converts unmethylated cytosines into uracils, while leaving 5-methylcytosines (5mC) unchanged. The methylation status at each CpG site is then inferred by comparing the ratio of methylated to unmethylated cytosines^[29]. After BS treatment, the DNA can be analyzed using either microarray-based platforms or next-generation sequencing (NGS) technologies.

1.2 DNA methylation arrays

Illumina’s Infinium methylation arrays are among the most commonly used platforms for DNA methylation analysis due to their relatively low cost, scalability, and ease of implementation. These arrays are designed to target specific CpG sites in the human genome using oligonucleotide probes that hybridize with BS-treated DNA. The hybridization signal is then measured to quantify methylation levels. Several versions of the Infinium arrays have been developed, including the Human Methylation 27 (27K)^[30], Human Methylation 450 (450K)^[31], and Human MethylationEPIC (EPIC) BeadChips^[32]. The latest version, the EPIC array, interrogates over 850 000 CpG sites. In this review, the EPIC array is used as a primary example to illustrate array-based methylation detection strategies.

1.3 DNA methylation sequencing

Sequencing-based DNA methylation detection offers high accuracy and broad genome coverage. This approach uses high-throughput sequencing to examine BS-converted DNA at single-base resolution. DNA methylation sequencing methods can be divided into 2 categories: Bulk-tissue sequencing and single-cell sequencing. For bulk tissues, reduced representation bisulfite sequencing (RRBS) and WGBS are most commonly used^[33-34]. RRBS is cost-effective and selectively captures CpG-rich regions, covering 1% to 5% of the genome. In contrast, WGBS is considered the gold standard, providing genome-wide, base-resolution data.

Single-cell DNA methylation sequencing enables detection of cell-type-specific methylation patterns, which may be masked in bulk-tissue analyses. This is particularly useful in complex tissues such as the brain, or in environments with heterogeneous cellular responses. Techniques include single-cell RRBS (scRRBS), single-cell WGBS (scWGBS), and single-nucleus methylcytosine sequencing (snmC-seq)^[35-37]. In spaceflight contexts-where environmental stressors such as microgravity or cosmic radiation may elicit distinct epigenetic responses across cell types-single-cell approaches offer a promising avenue for dissecting neuroimmune or neurovascular vulnerability. As WGBS remains the most comprehensive and widely accepted method for genome-wide methylation profiling, this review emphasizes WGBS as a representative sequencing-based platform.

2 Three phases in DNA methylation studies

DNA methylation studies typically progress through 3 critical phases: Study design, sample preparation and detection, and data analysis. Each phase presents unique methodological challenges that may significantly influence the validity, reproducibility, and biological interpretability of results-especially in the context of complex conditions such as NPDs and emerging environmental paradigms, including long-duration spaceflight. In addition to these 3 core phases, this section also introduces a workflow model for DNA methylation-based mental health monitoring in astronauts, illustrating how established methodologies can be adapted for space biomedical applications. Together, these subsections provide a structured framework and practical considerations for conducting genome-wide DNA methylation studies across terrestrial and spaceflight contexts.

2.1 Study design

Multiple biological, behavioral, and demographic factors-including sex, age, ancestry, body mass index (BMI), smoking, alcohol consumption, medication use, educational attainment, circadian rhythms, and sample size-can substantially influence the results of DNA methylation studies. Beyond these conventional variables, growing evidence^[2-6] points to the impact of extreme environments, particularly long-duration spaceflight, in reshaping the epigenome. Careful consideration and control of these variables during study design and participant enrollment are essential for minimizing bias, mitigating confounding, and improving the overall validity and reproducibility of findings across both terrestrial and space-based research settings.

2.1.1 Sex

Sex differences are well established in many NPDs and can influence DNA methylation signatures^[38-40]. For example, approximately two-thirds of AD patients are female, whereas around 80% of ASD patients are male^[41-42]. A transcriptomic analysis identified over 2 000 sex-biased genes linked to psychiatric risk, many of which are enriched in synaptic pathways^[39]. Given these differences, researchers should account for sex effects by matching case and control sex ratios or restricting analyses to a single sex to reduce confounding.

2.1.2 Age

Age is a well-known determinant of DNA methylation variation. Methylation patterns evolve across the lifespan, with global methylation levels generally declining with age^[43-44]. Certain CpG sites show strong correlations with chronological age and serve as epigenetic clocks for biological aging estimation^[45-48]. Therefore, age-matching between cases and controls is essential to minimize age-related confounding in NPD studies.

2.1.3 Ancestry

Ancestry influences DNA methylation through both genetic and environmental mechanisms^[49-50]. Ethnic differences in methylation profiles have been observed as early as birth, implicating inherited variation^[51]. As population structure may confound case-control analyses, especially in admixed cohorts, ancestry should be measured and included as a covariate or addressed via population-matched sampling.

2.1.4 BMI

BMI, a complex trait shaped by genetics, behavior, and environment, has been associated with methylation differences at numerous CpG sites^[52-53]. In NPDs cohorts, BMI may also reflect medication side effects or lifestyle differences^[53]. As such, it should be measured and adjusted for when assessing methylation outcomes.

2.1.5 Smoking

Smoking status is a robust modifier of DNA methylation, with over 1 000 CpG sites linked to tobacco exposure^[54-56]. Some methylation changes reverse after cessation, while others persist for decades^[54]. Smoking history should be carefully collected and included in statistical models to control for its widespread impact.

2.1.6 Alcohol consumption

Alcohol intake is another behavioral factor associated with differential methylation^[57]. In one study^[58], 144 CpG sites were correlated with alcohol use, and MZ twins discordant for drinking showed distinct methylation patterns. Although replication is needed, alcohol consumption should be recorded and considered during analysis.

2.1.7 Drug use

Pharmacological treatments commonly used in NPDs-such as antipsychotics or antidepressants-can induce both global and site-specific methylation changes^{[9, 59]}. These effects may also interact with disease-related biology. Accurate recording of medication type, dosage, and duration is critical to account for treatment-related confounding.

2.1.8 Educational attainment

Educational attainment has been linked to epigenetic variation, potentially reflecting broader life exposures such as nutrition, stress, or socioeconomic status. A population-level study^[60] identified 58 CpG sites associated with education, many tied to cognitive or health outcomes. Researchers should document education level and consider it as a socioeconomic proxy in analyses.

2.1.9 Circadian rhythms

Circadian disruption has been shown to alter methylation patterns. For instance, MZ twins with differing chronotypes exhibited epigenetic differences^[61], and studies in shift workers have confirmed this association^[62-63]. Standardizing sample collection time, or including sampling time as a covariate, can reduce variability related to circadian effects.

2.1.10 Sample size and statistical power

Sample size is a critical determinant of the accuracy and reliability of DNA methylation studies. Small sample sizes can amplify the effects of individual variability, reducing the generalizability of findings. Although MZ twin studies can detect DNA methylation differences, their results may not be applicable to larger cohorts^[64]. Therefore, increasing the sample size can help researchers identify more nuanced differences and enhance the study’s overall validity.

Statistical power is equally important for detecting meaningful differences in DNA methylation. For example, with the Illumina EPIC array, a sample size of approximately 1 000 participants (500 cases and 500 controls) is required to detect a 2% difference in methylation at most CpG sites with 100% power^[65]. Such a sample size enables the detection of subtle differences at most CpG sites. For WGBS, read depth is crucial for enhancing both accuracy and statistical power^[66]. Researchers are advised to aim for an effective sequencing depth of approximately 30× per sample to ensure accurate methylation detection^[67]. Both the EPIC array and WGBS offer statistical power calculation tools to guide study design, helping researchers determine the appropriate sample size needed to detect expected differences, maximize resource use, and minimize waste^[65-66].

2.1.11 Deep space environmental factors

Spaceflight introduces unique environmental stressors—including microgravity^{[2, 4]}, high-energy radiation^[2-3], circadian desynchrony^[5-6], confinement^[5-6], and psychosocial isolation^[5-6]―that can reshape DNA methylation landscapes. Studies from the NASA Twins Study^[6], Mars500^[5], and rodent spaceflight models^[2-3] have shown that these factors can trigger both transient and persistent methylation changes, particularly in brain and immune pathways. For instance, Impey, et al^[3] reported hippocampal 5-hydroxymethylcytosine (5hmC) changes following radiation exposure, while Hou, et al^[5] observed reversible methylation shifts under isolation. As human exploration pushes further into deep space, such factors must be incorporated into study design when investigating methylation biomarkers of neuropsychiatric vulnerability in space medicine contexts.

2.2 Sample preparation and detection

The sample preparation and detection phase is critical in DNA methylation studies, encompassing biospecimen collection, experimental procedures, and data acquisition. To ensure reliable and reproducible results, researchers must account for several technical factors that can introduce non-biological variation, including batch effects, positional effects in array-based platforms, and protocol-induced biases in WGBS library preparation. These considerations are particularly important in multi-site or longitudinal studies-including those conducted in extreme environments such as analog space habitats or actual spaceflight settings-where standardization and contamination control are essential for meaningful comparisons.

2.2.1 Batch effects

Batch effects are a common source of technical bias in DNA methylation studies and can confound biological interpretations if not properly controlled^[68]. These effects may arise from differences in sample collection times, processing personnel, reagent lots, or instrumentation^[69-70]. As batch effects can mimic or obscure true biological signals, especially in case-control comparisons, proactive mitigation strategies are essential. These include consistent experimental protocols, randomized sample allocation across batches, and statistical correction using tools such as ComBat during data analysis. In space biomedical research—where samples may be collected at different mission timepoints or across isolated environments—batch normalization becomes particularly crucial.

2.2.2 Positional effects

Positional effects occur when the physical location of a sample on a methylation array influences signal intensity or methylation estimates^[71-72]. These biases are inherent to the array design and can result in systematic variation unrelated to biological factors. To minimize such effects, samples should be spatially balanced across chips (e.g., evenly distributing cases and controls), and randomization of sample placement is recommended^[71]. Computational adjustment methods can also be applied post hoc. In studies involving in-flight sample processing or transport delays, maintaining strict sample layout protocols is essential for downstream data comparability.

2.2.3 Protocol selection and bias in WGBS library preparation

In WGBS, the choice of library preparation protocol can significantly influence data quality. Library construction generally involves 4 steps: DNA fragmentation, adapter ligation, BS conversion, and PCR amplification. There are 2 main strategies based on the order of these steps: pre-BS and post-BS library preparation^[73-74].

In the pre-BS protocol^[73], adapters are ligated before BS conversion. However, BS treatment leads to extensive DNA degradation, reducing fragment yield and requiring higher input amounts. In contrast, the post-BS protocol^[74] performs adapter ligation after BS conversion, preserving more DNA and allowing for high-quality libraries with lower DNA input. For studies with limited sample material-such as astronaut-derived blood samples or simulated microgravity models-the post-BS protocol is generally preferred for its efficiency and reduced DNA requirements.

Biases introduced during BS conversion and PCR amplification can further affect methylation estimates. Incomplete conversion of unmethylated cytosines or preferential amplification may result in artifactual methylation patterns^[75]. Although these biases cannot be fully eliminated, researchers are advised to use commercial kits with conversion efficiencies ≥99% and minimize PCR cycles to reduce amplification artifacts. Such precautions are especially important in low-input or precious samples derived from spaceflight studies, where sample availability and integrity are limiting factors.

2.3 Data analysis

DNA methylation data analysis typically involves multiple steps, from raw signal pre-processing to downstream statistical and functional interpretation. To ensure accurate and reproducible results, workflows should include quality control, normalization, batch-effect correction, and removal of unreliable probes. Although the Illumina MethylationEPIC array and WGBS follow similar analytical principles, they differ in data structure, pre-processing complexity, and computational tools. Table 1 summarizes the major features of the 2 platforms in terms of pre-processing, quality control, and downstream analysis tools.

2.3.1 Pre-processing

2.3.1.1 Quality control

EPIC and WGBS follow similar principles for pre-processing but differ in execution. EPIC arrays typically employ the ChAMP pipeline for Illumina Intensity Data (IDAT) loading, probe and sample filtering, and β-value normalization, while WGBS involves adapter trimming, quality assessment, alignment, and BS conversion efficiency checks^{[31-32, 76-86]}. The shared goal is to minimize technical noise and improve comparability across samples. This is particularly important in spaceflight or analog studies because sample degradation, low input, and inter-mission batch effects are more likely, making robust pre-processing and quality control especially critical.

2.3.1.2 Confounding factor correction

Confounding variables identified during study design must also be addressed at the analysis stage. This includes: Correction for batch and positional effects using ComBat (in the sva package); adjustment for cell type heterogeneity through deconvolution methods; detection of unmeasured confounders using num.sv and svaseq; linear modeling to account for both known and latent variables^[87]. These steps help isolate true biological signals, a critical requirement when comparing methylation across heterogeneous or extreme conditions such as terrestrial vs spaceflight cohorts.

2.3.2 Downstream analysis

2.3.2.1 Differential DNA methylation analysis

Differentially methylated positions (DMPs) and differentially methylated regions (DMRs) can be identified using platform-specific tools summarized in Table 1^[88-95]. These approaches are essential for uncovering disease-related epigenetic changes in NPD studies and can also be extended to spaceflight-induced methylation alterations, especially when evaluating mission-phase-specific shifts.

2.3.2.2 Epigenome-Wide Association Studies and Methylation-Wide Association Studies

Epigenome-Wide Association Studies (EWAS) and Methylation-Wide Association Studies (MWAS) are large-scale analytic strategies used to relate methylation profiles with phenotypes or disease outcomes^[96-99]. EWAS^{[96, 99]} broadly examine methylation variation in association with exposures (e.g., stress, smoking, radiation) or traits; MWAS^[97-98] focuse more directly on identifying methylation signatures associated with specific diseases, such as SCZ or BD. These approaches are increasingly valuable in longitudinal monitoring of individuals exposed to extreme environments, such as astronauts undergoing psychological and physiological stress during long-duration missions.

2.3.2.3 Annotation and functional analysis

Methylated loci can be annotated using databases such as ENCODE or the UCSC Genome Browser to determine their genomic context (e.g., promoters, enhancers, CpG islands, repetitive elements). Functional enrichment can then be explored via tools such as Gene Ontology (GO), Gene Set Enrichment Analysis (GSEA), and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway mapping^[100-102]. These analyses help elucidate the biological relevance of methylation alterations. In the context of NPDs^[8-11], enriched pathways often involve neurodevelopment, synaptic transmission, immune signaling, or stress response-mechanisms also perturbed by spaceflight stressors. Integrating functional annotation with exposure metadata (e.g., mission duration, radiation dose) provides a mechanistic framework for interpreting space-associated neuroepigenetic changes.

2.4 Workflow for DNA methylation-based neuropsychiatric health monitoring in astronauts

DNA methylation, as an environmentally sensitive, reversible, and peripherally accessible epigenetic marker, offers unique potential as an early and non-invasive indicator of neuropsychiatric health risks in astronauts. Unlike traditional psychological or hormonal assessments, DNA methylation can capture long-term or cumulative stress effects at the molecular level, providing a new entry point for multi-modal health surveillance.

2.4.1 Pre-flight baseline assessment

Peripheral samples such as whole blood, dried blood spots (DBS), or saliva should be collected before mission launch to establish individualized methylation profiles. Priority targets include pathways linked to neuropsychiatric risk^[8-11]―HPA-axis stress regulation, circadian rhythm genes, neuroplasticity and neurotransmission, and immune-inflammatory and oxidative stress pathways. Multiple baseline time points can estimate intra-individual variability, while combining methylation data with psychological scales, cognitive tests, and sleep/physiological parameters helps identify high-risk individuals and define personalized thresholds.

2.4.2 In-flight longitudinal monitoring

During missions, low-burden, small-volume sampling (e.g., DBS, stabilized tubes, or saliva) at regular intervals can track dynamic changes in key CpG sites, module scores, or methylation profile scores (MPS)^[26] relative to individual baselines. Integration with psychological questionnaires, cognitive tasks, and physiological sensor data supports a multi-modal longitudinal monitoring framework. Particular attention should be paid to methylation markers^[8-11] of stress, circadian disruption, neuroplasticity, and immune-inflammatory processes to enable early detection of adverse trends and guide countermeasures such as circadian adjustment, workload modification, and psychological support.

2.4.3 Post-flight recovery and long-term evaluation

After landing, continued monitoring of key methylation indicators^[22-23] can assess reversibility and long-term changes, linking these patterns to cognitive recovery, psychological adaptation, and immune homeostasis. Higher-resolution approaches (e.g., WGBS or single-cell methylation analysis) may be used for in-depth validation to refine health management and inform future mission planning.

Limited sample volumes, storage constraints, ground-based analysis requirements, and the need to balance sampling frequency with crew workload must all be addressed. Longitudinal, within-subject analyses anchored to pre-flight baselines are recommended to improve sensitivity and specificity. Collectively, DNA methylation represents a valuable adjunct to astronaut neuropsychiatric health surveillance, providing a molecular foundation for risk identification and intervention during long-duration space missions.

3 Choosing the appropriate DNA methylation detection method

The selection of an appropriate DNA methylation detection platform depends on the specific objectives, sample availability, and technical requirements of a given study. In NPDs research, 2 mainstream approaches―EPIC array and WGBS―are commonly used to compare patients and controls, detect DMPs or DMRs, and develop methylation-based risk metrics^[103]. Recently, MPS, conceptually analogous to polygenic scores (PGS), have gained traction as quantitative markers of epigenetic risk^[25-27].

3.1 EPIC array

The EPIC array is widely used in NPDs research due to its affordability, scalability, and ease of analysis. It provides single-CpG resolution for over 850 000 targeted CpG sites and allows multiplexing of up to 8 samples per chip. Bioinformatic pipelines such as ChAMP and minfi facilitate rapid data processing, making the platform accessible to research groups with limited computational infrastructure.

In addition to DMP and DMR detection, the EPIC array enables methylation quantitative trait locus (meQTL) studies. Public reference datasets from blood and brain samples support cross-study harmonization^[104-105]. For instance, Li, et al^[106] used EPIC arrays to analyze blood samples from 469 first-episode schizophrenia (FESZ) patients and 476 matched controls in a Han Chinese cohort, identifying DMPs associated with neuronal excitability and neuro-development.

In space biomedical applications, the EPIC array offers practical advantages for population-level screening, including cost-effectiveness, high throughput, and minimal DNA input requirements. These features are particularly valuable in large analog studies or in-flight sample collection scenarios where logistical constraints limit sequencing-based approaches.

3.2 WGBS

WGBS remains the gold standard for comprehensive DNA methylation profiling, offering single-base resolution across nearly all CpG sites. It facilitates the discovery of novel regulatory elements, characterization of non-coding regions, and in-depth mapping of the epigenetic landscape. However, WGBS is both cost-intensive and computationally demanding. The library preparation process is laborious, and downstream analysis requires substantial bioinformatic expertise^[107]. Recent technological advances, such as the release of the Telomere-to-Telomere CHM13 (T2T-CHM13) reference genome, have improved read alignment accuracy and genome coverage, mitigating some of these limitations^[108].

WGBS is particularly useful for genome-wide meQTL mapping, mechanistic investigations, and biomarker discovery^[109]. It is also compatible with single-cell techniques such as scWGBS, which allow for the resolution of cell-type-specific methylation signatures within complex tissues^[110]. For example, Zhu, et al^[111] identified 134 DMRs associated with autism ASD using WGBS on placenta samples, while Mendizabal, et al^[112] applied WGBS to postmortem brain tissues from individuals with SCZ and reported that intercellular variability exceeded case-control differences.

In the context of deep space research, WGBS is uniquely suited to capturing both localized and global epigenetic responses to environmental stressors such as microgravity, cosmic radiation, and chronic isolation. Its high resolution makes it an essential tool for elucidating neuroimmune mechanisms underlying spaceflight, induced psychiatric vulnerability, and for developing molecular biomarkers to support mental health monitoring during long-duration missions.

3.3 Comparison and insights

The EPIC array and WGBS serve complementary roles in DNA methylation research. The EPIC array is well suited for large-scale epidemiological studies, biomarker screening, and resource-limited projects, offering cost-efficiency, multiplexing capability, and standardized data workflows. WGBS, by contrast, provides exhaustive genome-wide coverage and single-base resolution, making it ideal for mechanistic investigations, meQTL mapping, and discovery of non-coding regulatory elements^[113-114].

In the context of space medicine, these distinctions remain critical. The EPIC array is particularly advantageous for screening astronauts before and after missions or for deployment in analog environments, where throughput and simplicity are priorities. WGBS, meanwhile, is essential for hypothesis-driven studies aiming to uncover subtle or system-wide epigenetic effects of microgravity, radiation, or psychological isolation. Single-cell WGBS further enables resolution of cell-type-specific epigenetic changes in neural, immune, or vascular tissues-key systems responsive to spaceflight stressors^[5-6]. Ultimately, platform selection should be guided by the study’s goals, sample constraints, and analytical demands. In many cases, an integrated strategy leveraging both EPIC and WGBS may provide the optimal balance between practical implementation and scientific depth, especially when translating bench findings into operational tools for long-duration space missions.

While current platforms such as EPIC and WGBS provide robust frameworks for methylation profiling, several technical, environmental, and interpretive challenges remain-particularly when extending such studies to extreme settings like spaceflight.

4 Emerging challenges in neuropsychiatric methylation research under Earth and space conditions

DNA methylation studies^{[11-12, 115-122]} in SCZ, BD, major depressive disorder, AD, and PD highlight shared epigenetic pathways involving stress regulation (NR3C1, FKBP5/HPA axis), neuroplasticity (BDNF, RELN, GAD1, DUSP22), immune and oxidative stress (ANK1, RHBDF2, SLC7A11, DJ-1, CYP2E1), and neurotransmitter or circadian control (SLC6A4, DAT1, SNCA). These systems overlap with spaceflight stressors such as chronic isolation, radiation, microgravity and circadian disruption, suggesting their potential as molecular targets for astronaut neuropsychiatric monitoring. Prolonged mission stress and disrupted circadian rhythms may remodel stress and plasticity pathways; radiation and microgravity may amplify immune-related methylation changes; and cumulative stressors may destabilize SNCA-DNMT1 networks and alter circadian and neurotransmitter genes, affecting cognition and emotion. Integrating such methylation signatures with psychological, physiological, and behavioral data could support multi-modal, longitudinal, individualized risk assessment during long-duration missions (Table 2).

NPDs are complex conditions shaped by genetic predisposition and environmental exposures. Among the tools for dissecting this complexity, meQTL analysis^[104-105] has gained prominence for exploring how genetic variants regulate DNA methylation patterns and contribute to disease risk. However, most current meQTL studies rely on array-based platforms such as the EPIC array, which offer limited genomic coverage. WGBS^[123], by contrast, enables more comprehensive genome-wide meQTL mapping across both coding and non-coding regions, offering deeper insights into regulatory variation.

Environmental factors-ranging from chronic psychosocial stress to air pollution and dietary variation-can dynamically shape the methylome, often in a tissue-specific or cell-type-specific manner^[124-126]. Recent attention has turned to spaceflight-induced exposures, including microgravity, high-energy radiation, circadian misalignment, and isolation. These extreme conditions have been shown to induce stable methylation changes, particularly in neuroimmune pathways, underscoring the need for epigenetic models that can incorporate both terrestrial and extraterrestrial stressors^{[2-3, 5-6]}. Accounting for such complex and chronic exposures will be essential to improve the reliability, cross-context interpretability, and translational relevance of neuroepigenetic data.

The human brain remains the primary site of dysfunction in NPDs, but direct sampling is typically not feasible. Peripheral blood is thus commonly used as a surrogate tissue. Establishing robust correlations between methylation profiles in brain and blood, and identifying consistent cross-tissue biomarkers, is key to improving the clinical applicability of peripheral measurements^[127-128]. In this context, deep space missions further restrict invasive procedures, emphasizing the value of minimally invasive, blood-based biomarkers for real-time mental health surveillance. Recent progress^[129-131] in induced pluripotent stem cell (iPSC) technologies and brain organoid models has opened new avenues for modeling neurodevelopmental epigenetics. These tools are especially promising for simulating in vivo responses to environmental perturbations, including those relevant to spaceflight, and for testing the reversibility of epigenetic states in a controlled setting.

Rapid technological innovation continues to expand the utility of DNA methylation research. Machine learning algorithms have been applied to large methylation datasets to identify stable DMPs or DMRs associated with disease onset and progression^[132-133]. For example, Gunasekara, et al^[133] analyzed blood DNA methylation profiles from over 800 SCZ patients and controls using a sparse partial least squares discriminant analysis (SPLS-DA) framework focused on systemic interindividual epigenetic variation (CoRSIVs). Their approach achieved approximately 80% positive predictive value in an independent validation cohort and revealed strong blood-brain concordance at key loci, underscoring the feasibility of peripheral samples for central nervous system-related epigenetic research. MPS, which quantify disease risk based on epigenetic patterns, provide a potentially reversible alternative to static PGS. However, MPS performance is influenced by temporal variation in methylation and requires validation in well-characterized, longitudinal cohorts^[26]. These validation efforts should increasingly include participants exposed to extreme environments such as long-duration spaceflight, to assess how chronic physiological stress alters methylation-based predictive models.

Advances in genomic reference resources and sequencing technologies further enhance research capabilities. The release of the telomere-to-telomere T2T-CHM13 human reference genome, which spans over 32.29 million CpG sites compared to approximately 29.2 million in GRCh38, has improved WGBS mapping accuracy and coverage, allowing better resolution of regulatory elements^[108]. In parallel, novel detection platforms, such as high-efficiency single-cell methylation sequencing and BS-free technologies (e.g., Oxford nanopore), are being developed^[134-137]. These approaches reduce conversion bias and enable real-time, portable methylation profiling-capabilities particularly well suited to spaceflight scenarios, including in-orbit or planetary base diagnostics where minimal equipment and rapid turnaround are required.

Looking forward, DNA methylation research in NPDs is expected to evolve toward increasingly large-scale, high-resolution, and multi-context frameworks. As sequencing costs decline and computational infrastructure improves, WGBS will likely become more widely adopted for both Earth-based and space-integrated studies^{[111, 113-114]}. These advancements will not only accelerate discovery of epigenetic mechanisms underlying psychiatric disease but also inform personalized diagnostic strategies and countermeasures to maintain cognitive and emotional resilience in space. Ultimately, integrating DNA methylation technologies into the fabric of space medicine offers a new frontier for real-time mental health assessment, molecular risk stratification, and long-term human adaptation beyond Earth.

参考文献

[1] Krittanawong C, Singh NK, Scheuring RA, et al. Human health during space travel: state-of-the-art review[J]. Cells, 2022, 12(1): 40.

[2] Overbey EG, Paul AM, da Silveira WA, et al. Mice exposed to combined chronic low-dose irradiation and modeled microgravity develop long-term neurological sequelae[J]. Int J Mol Sci, 2019, 20(17): 4094.

[3] Impey S, Pelz C, Riparip LK, et al. Postsynaptic density radiation signature following space irradiation[J]. Front Physiol, 2023, 14: 1215535.

[4] Chowdhury B, Seetharam A, Wang ZP, et al. A study of alterations in DNA epigenetic modifications (5mC and 5hmC) and gene expression influenced by simulated microgravity in human lymphoblastoid cells[J/OL]. PLoS One, 2016, 11(1): e0147514[2025-05-12].

[5] Hou F, Zhou X, Zhou S, et al. DNA methylation dynamics associated with long-term isolation of simulated space travel [J]. iScience, 2022, 25(6): 104493.

[6] Garrett-Bakelman FE, Darshi M, Green SJ, et al. The NASA Twins Study: a multidimensional analysis of a year-long human spaceflight[J/OL]. Science, 2019, 364(6436): eaau8650[2025-05-12].

[7] Feng J, Fan G. The role of DNA methylation in the central nervous system and neuropsychiatric disorders[J]. Int Rev Neurobiol, 2009, 89: 67-84.

[8] Ai S, Shen L, Guo J, et al. DNA methylation as a biomarker for neuropsychiatric diseases[J]. Int J Neurosci, 2012, 122(4): 165-176.

[9] Liu C, Jiao C, Wang K, et al. DNA methylation and psychiatric disorders[M]// Epigenetics and Psychiatric Disease. Amsterdam: Elsevier, 2018: 175-232.

[10] Kuehner JN, Bruggeman EC, Wen ZX, et al. Epigenetic regulations in neuropsychiatric disorders[J]. Front Genet, 2019, 10: 268.

[11] Shirvani-Farsani Z, Maloum Z, Bagheri-Hosseinabadi Z, et al. DNA methylation signature as a biomarker of major neuropsychiatric disorders[J]. J Psychiatr Res, 2021, 141: 34-49.

[12] Rasmi Y, Shokati A, Hassan A, et al. The role of DNA methylation in progression of neurological disorders and neurodegenerative diseases as well as the prospect of using DNA methylation inhibitors as therapeutic agents for such disorders[J]. IBRO Neurosci Rep, 2022, 14: 28-37.

[13] Uher R, Zwicker A. Etiology in psychiatry: embracing the reality of poly-gene-environmental causation of mental illness[J]. World Psychiatry, 2017, 16(2): 121-129.

[14] Dunn AR, O’Connell KMS, Kaczorowski CC. Gene-by-environment interactions in Alzheimer’s disease and Parkinson’s disease[J]. Neurosci Biobehav Rev, 2019, 103: 73-80.

[15] Robinson N, Bergen SE. Environmental risk factors for schizophrenia and bipolar disorder and their relationship to genetic risk: current knowledge and future directions[J]. Front Genet, 2021, 12: 686666.

[16] Hilker R, Helenius D, Fagerlund B, et al. Heritability of schizophrenia and schizophrenia spectrum based on the nationwide Danish twin register[J]. Biol Psychiatry, 2018, 83(6): 492-498.

[17] Moore LD, Le T, Fan GP. DNA methylation and its basic function[J]. Neuropsychopharmacology, 2013, 38(1): 23-38.

[18] Tahiliani M, Koh KP, Shen Y, et al. Conversion of 5-methylcytosine to 5-hydroxymethylcytosine in mammalian DNA by MLL partner TET1[J]. Science, 2009, 324(5929): 930-935.

[19] Bird AP. CpG-rich islands and the function of DNA methylation[J]. Nature, 1986, 321(6067): 209-213.

[20] JANITZ K, JANITZ M. Chapter 12 - Assessing Epigenetic Information[M/OL]//TOLLEFSBOL T. Handbook of Epigenetics. San Diego: Academic Press, 2011: 173-181 [2023-03-09].

[21] Jones PA. Functions of DNA methylation: islands, start sites, gene bodies and beyond[J]. Nat Rev Genet, 2012, 13(7): 484-492.

[22] Mayeux R. Biomarkers: potential uses and limitations[J]. NeuroRx, 2004, 1(2): 182-188.

[23] Naylor S. Biomarkers: current perspectives and future prospects[J]. Expert Rev Mol Diagn, 2003, 3(5): 525-529.

[24] Mikeska T, Craig JM. DNA methylation biomarkers: cancer and beyond[J]. Genes, 2014, 5(3): 821-864.

[25] Nabais MF, Laws SM, Lin T, et al. Meta-analysis of genome-wide DNA methylation identifies shared associations across neurodegenerative disorders[J]. Genome Biol, 2021, 22(1): 90.

[26] Nabais MF, Gadd DA, Hannon E, et al. An overview of DNA methylation-derived trait score methods and applications[J]. Genome Biol, 2023, 24(1): 28.

[27] Ohi K, Shimada M, Soda M, et al. Genome-wide DNA methylation risk scores for schizophrenia derived from blood and brain tissues further explain the genetic risk in patients stratified by polygenic risk scores for schizophrenia and bipolar disorder[J/OL]. BMJ Ment Health, 2024, 27(1): e300936[2025-05-12].

[28] Yong WS, Hsu FM, Chen PY. Profiling genome-wide DNA methylation[J]. Epigenetics Chromatin, 2016, 9: 26.

[29] Frommer M, McDonald LE, Millar DS, et al. A genomic sequencing protocol that yields a positive display of 5-methylcytosine residues in individual DNA strands[J]. Proc Natl Acad Sci USA, 1992, 89(5): 1827-1831.

[30] Bibikova M, Le J, Barnes B, et al. Genome-wide DNA methylation profiling using Infinium® assay[J]. Epigenomics, 2009, 1(1): 177-200.

[31] Bibikova M, Barnes B, Tsan C, et al. High density DNA methylation array with single CpG site resolution[J]. Genomics, 2011, 98(4): 288-295.

[32] Pidsley R, Zotenko E, Peters TJ, et al. Critical evaluation of the Illumina MethylationEPIC BeadChip microarray for whole-genome DNA methylation profiling[J]. Genome Biol, 2016, 17(1): 208.

[33] Meissner A, Gnirke A, Bell GW, et al. Reduced representation bisulfite sequencing for comparative high-resolution DNA methylation analysis[J]. Nucleic Acids Res, 2005, 33(18): 5868-5877.

[34] Lister R, O’Malley RC, Tonti-Filippini J, et al. Highly integrated single-base resolution maps of the epigenome in Arabidopsis [J]. Cell, 2008, 133(3): 523-536.

[35] Guo H, Zhu P, Wu X, et al. Single-cell methylome landscapes of mouse embryonic stem cells and early embryos analyzed using reduced representation bisulfite sequencing[J]. Genome Res, 2013, 23(12): 2126-2135.

[36] Farlik M, Sheffield NC, Nuzzo A, et al. Single-cell DNA methylome sequencing and bioinformatic inference of epigenomic cell-state dynamics[J]. Cell Rep, 2015, 10(8): 1386-1397.

[37] Luo C, Keown CL, Kurihara L, et al. Single-cell methylomes identify neuronal subtypes and regulatory elements in mammalian cortex[J]. Science, 2017, 357(6351): 600-604.

[38] Xu H, Wang F, Liu Y, et al. Sex-biased methylome and transcriptome in human prefrontal cortex[J]. Hum Mol Genet, 2014, 23(5): 1260-1270.

[39] Xia Y, Dai R, Wang K, et al. Sex-differential DNA methylation and associated regulation networks in human brain implicated in the sex-biased risks of psychiatric disorders[J]. Mol Psychiatry, 2021, 26(3): 835-848.

[40] Zhang L, Young JI, Gomez L, et al. Sex-specific DNA methylation differences in Alzheimer’s disease pathology[J]. Acta Neuropathol Commun, 2021, 9(1): 77.

[41] Hebert LE, Weuve J, Scherr PA, et al. Alzheimer disease in the United States (2010-2050) estimated using the 2010 census[J]. Neurology, 2013, 80(19): 1778-1783.

[42] Werling DM, Geschwind DH. Sex differences in autism spectrum disorders[J]. Curr Opin Neurol, 2013, 26(2): 146-153.

[43] Jones MJ, Goodman SJ, Kobor MS. DNA methylation and healthy human aging[J]. Aging Cell, 2015, 14(6): 924-932.

[44] Field AE, Robertson NA, Wang TN, et al. DNA methylation clocks in aging: categories, causes, and consequences[J]. Mol Cell, 2018, 71(6): 882-895.

[45] Horvath S. DNA methylation age of human tissues and cell types[J]. Genome Biol, 2013, 14(10): R115.

[46] Zubakov D, Liu F, Kokmeijer I, et al. Human age estimation from blood using mRNA, DNA methylation, DNA rearrangement, and telomere length[J]. Forensic Sci Int Genet, 2016, 24: 33-43.

[47] Naue J, Hoefsloot HCJ, Mook ORF, et al. Chronological age prediction based on DNA methylation: Massive parallel sequencing and random forest regression[J]. Forensic Sci Int Genet, 2017, 31: 19-28.

[48] Horvath S, Raj K. DNA methylation-based biomarkers and the epigenetic clock theory of ageing[J]. Nat Rev Genet, 2018, 19(6): 371-384.

[49] Xia YY, Ding YB, Liu XQ, et al. Racial/ethnic disparities in human DNA methylation[J]. Biochim Biophys Acta BBA Rev Cancer, 2014, 1846(1): 258-262.

[50] Song MA, Seffernick AE, Archer KJ, et al. Race/ethnicity-associated blood DNA methylation differences between Japanese and European American women: an exploratory study[J]. Clin Epigenetics, 2021, 13(1): 188.

[51] Adkins RM, Krushkal J, Tylavsky FA, et al. Racial differences in gene-specific DNA methylation levels are present at birth[J]. Birth Defects Res A Clin Mol Teratol, 2011, 91(8): 728-736.

[52] Dick KJ, Nelson CP, Tsaprouni L, et al. DNA methylation and body-mass index: a genome-wide analysis[J]. Lancet, 2014, 383(9933): 1990-1998.

[53] Reed ZE, Suderman MJ, Relton CL, et al. The association of DNA methylation with body mass index: distinguishing between predictors and biomarkers[J]. Clin Epigenetics, 2020, 12(1): 50.

[54] Gao X, Jia M, Zhang Y, et al. DNA methylation changes of whole blood cells in response to active smoking exposure in adults: a systematic review of DNA methylation studies[J]. Clin Epigenetics, 2015, 7: 113.

[55] Ambatipudi S, Cuenin C, Hernandez-Vargas H, et al. Tobacco smoking-associated genome-wide DNA methylation changes in the EPIC study[J]. Epigenomics, 2016, 8(5): 599-618.

[56] Saint-André V, Charbit B, Biton A, et al. Smoking changes adaptive immunity with persistent effects[J]. Nature, 2024, 626(8000): 827-835.

[57] Liu C, Marioni RE, Hedman ÅK, et al. A DNA methylation biomarker of alcohol consumption[J]. Mol Psychiatry, 2018, 23(2): 422-433.

[58] Meng L, Qin XY, Peng HX, et al. Genome-wide associations between alcohol consumption and blood DNA methylation: evidence from twin study[J]. Epigenomics, 2021, 13(12): 939-951.

[59] Reynolds GP, Fachim HA. Does DNA methylation influence the effects of psychiatric drugs?[J]. Epigenomics, 2016, 8(3): 309-312.

[60] van Dongen J, Bonder MJ, Dekkers KF, et al. DNA methylation signatures of educational attainment[J]. NPJ Sci Learn, 2018, 3: 7.

[61] Wong CCY, Parsons MJ, Lester KJ, et al. Epigenome-wide DNA methylation analysis of monozygotic twins discordant for diurnal preference[J]. Twin Res Hum Genet, 2015, 18(6): 662-669.

[62] Bhatti P, Zhang Y, Song X, et al. Nightshift work and genome-wide DNA methylation[J]. Chronobiol Int, 2015, 32(1): 103-112.

[63] Adams CD, Jordahl KM, Copeland W, et al. Nightshift work, chronotype, and genome-wide DNA methylation in blood[J]. Epigenetics, 2017, 12(10): 833-840.

[64] Bell JT, Spector TD. DNA methylation studies using twins: what are they telling us?[J]. Genome Biol, 2012, 13(10): 172.

[65] Mansell G, Gorrie-Stone TJ, Bao Y, et al. Guidance for DNA methylation studies: statistical insights from the Illumina EPIC array[J]. BMC Genomics, 2019, 20(1): 366.

[66] Seiler Vellame D, Castanho I, Dahir A, et al. Characterizing the properties of bisulfite sequencing data: maximizing power and sensitivity to identify between-group differences in DNA methylation[J]. BMC Genomics, 2021, 22(1): 446.

[67] Ziller MJ, Hansen KD, Meissner A, et al. Coverage recommendations for methylation analysis by whole-genome bisulfite sequencing[J]. Nat Methods, 2015, 12(3): 230-232, 1pfollowing232.

[68] Price EM, Robinson WP. Adjusting for batch effects in DNA methylation microarray data, a lesson learned[J]. Front Genet, 2018, 9: 83.

[69] Leek JT, Scharpf RB, Bravo HC, et al. Tackling the widespread and critical impact of batch effects in high-throughput data[J]. Nat Rev Genet, 2010, 11(10): 733-739.

[70] Chen C, Grennan K, Badner J, et al. Removing batch effects in analysis of expression microarray data: an evaluation of six batch adjustment methods[J/OL]. PLoS One, 2011, 6(2): e17238[2025-05-12].

[71] Verdugo RA, Deschepper CF, Muñoz G, et al. Importance of randomization in microarray experimental designs with Illumina platforms[J]. Nucleic Acids Res, 2009, 37(17): 5610-5618.

[72] Jiao C, Zhang CL, Dai RJ, et al. Positional effects revealed in Illumina methylation array and the impact on analysis[J]. Epigenomics, 2018, 10(5): 643-659.

[73] Cokus SJ, Feng SH, Zhang XY, et al. Shotgun bisulphite sequencing of the Arabidopsis genome reveals DNA methylation patterning[J]. Nature, 2008, 452(7184): 215-219.

[74] Miura F, Enomoto Y, Dairiki R, et al. Amplification-free whole-genome bisulfite sequencing by post-bisulfite adaptor tagging[J/OL]. Nucleic Acids Res, 2012, 40(17): e136[2025-05-12].

[75] Olova N, Krueger F, Andrews S, et al. Comparison of whole-genome bisulfite sequencing library preparation strategies identifies sources of biases affecting DNA methylation data[J]. Genome Biol, 2018, 19(1): 33.

[76] Morris TJ, Butcher LM, Feber A, et al. ChAMP: 450k chip analysis methylation pipeline[J]. Bioinformatics, 2014, 30(3): 428-430.

[77] Tian Y, Morris TJ, Webster AP, et al. ChAMP: updated methylation analysis pipeline for Illumina BeadChips[J]. Bioinformatics, 2017, 33(24): 3982-3984.

[78] Aryee MJ, Jaffe AE, Corrada-Bravo H, et al. Minfi: a flexible and comprehensive Bioconductor package for the analysis of Infinium DNA methylation microarrays[J]. Bioinformatics, 2014, 30(10): 1363-1369.

[79] Fortin JP, Triche TJ, Hansen KD. Preprocessing, normalization and integration of the Illumina HumanMethylationEPIC array with minfi[J]. Bioinformatics, 2017, 33(4): 558-560.

[80] Teschendorff AE, Marabita F, Lechner M, et al. A beta-mixture quantile normalization method for correcting probe design bias in Illumina Infinium 450 k DNA methylation data[J]. Bioinformatics, 2013, 29(2): 189-196.

[81] Fortin JP, Labbe A, Lemire M, et al. Functional normalization of 450 k methylation array data improves replication in large cancer studies[J]. Genome Biol, 2014, 15(12): 503.

[82] Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads[J]. EMBnet J, 2011, 17(1): 10.

[83] Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data[J]. Bioinformatics, 2014, 30(15): 2114-2120.

[84] Babraham Bioinformatics―FastQC A quality control tool for high throughput sequence data[EB/OL]. [2022-07-30].

[85] Krueger F, Andrews SR. Bismark: a flexible aligner and methylation caller for Bisulfite-Seq applications[J]. Bioinformatics, 2011, 27(11): 1571-1572.

[86] Guo H, Zhu P, Yan L, et al. The DNA methylation landscape of human early embryos[J]. Nature, 2014, 511(7511): 606-610.

[87] Johnson WE, Li C, Rabinovic A. Adjusting batch effects in microarray expression data using empirical Bayes methods[J]. Biostatistics, 2007, 8(1): 118-127.

[88] Wettenhall JM, Smyth GK. limmaGUI: a graphical user interface for linear modeling of microarray data[J]. Bioinformatics, 2004, 20(18): 3705-3706.

[89] Smyth GK. Linear models and empirical Bayes methods for assessing differential expression in microarray experiments[J]. Stat Appl Genet Mol Biol, 2004, 3: Article3.

[90] Hansen KD, Timp W, Bravo HC, et al. Increased methylation variation in epigenetic domains across cancer types[J]. Nat Genet, 2011, 43(8): 768-775.

[91] Timp W, Bravo HC, McDonald OG, et al. Large hypomethylated blocks as a universal defining epigenetic alteration in human solid tumors[J]. Genome Med, 2014, 6(8): 61.

[92] Akalin A, Kormaksson M, Li S, et al. methylKit: a comprehensive R package for the analysis of genome-wide DNA methylation profiles[J]. Genome Biol, 2012, 13(10): R87.

[93] Hansen KD, Langmead B, Irizarry RA. BSmooth: from whole genome bisulfite sequencing reads to differentially methylated regions[J]. Genome Biol, 2012, 13(10): R83.

[94] Feng H, Conneely KN, Wu H. A Bayesian hierarchical model to detect differentially methylated loci from single nucleotide resolution sequencing data[J/OL]. Nucleic Acids Res, 2014, 42(8): e69[2025-05-12].

[95] Dolzhenko E, Smith AD. Using beta-binomial regression for high-precision differential methylation analysis in multifactor whole-genome bisulfite sequencing experiments[J]. BMC Bioinformatics, 2014, 15: 215.

[96] Campagna MP, Xavier A, Lechner-Scott J, et al. Epigenome-wide association studies: current knowledge, strategies and recommendations[J]. Clin Epigenetics, 2021, 13(1): 214.

[97] Shabalin AA, Hattab MW, Clark SL, et al. RaMWAS: fast methylome-wide association study pipeline for enrichment platforms[J]. Bioinformatics, 2018, 34(13): 2283-2285.

[98] Hop PJ, Zwamborn RAJ, Hannon E, et al. Genome-wide study of DNA methylation shows alterations in metabolic, inflammatory, and cholesterol pathways in ALS[J/OL]. Sci Transl Med, 2022, 14(633): eabj0264[2025-01-25].

[99] Silva TC, Young JI, Zhang LY, et al. Cross-tissue analysis of blood and brain epigenome-wide association studies in Alzheimer’s disease[J]. Nat Commun, 2022, 13: 4852.

[100] Subramanian A, Tamayo P, Mootha VK, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles[J]. Proc Natl Acad Sci U S A, 2005, 102(43): 15545-15550.

[101] Consortium The Gene Ontology. Gene ontology consortium: going forward[J]. Nucleic Acids Res, 2015, 43(D1): D1049-D1056.

[102] Kanehisa M, Sato Y, Kawashima M, et al. KEGG as a reference resource for gene and protein annotation[J]. Nucleic Acids Res, 2016, 44(D1): D457-D462.

[103] Kular L, Kular S. Epigenetics applied to psychiatry: Clinical opportunities and future challenges[J]. Psychiatry Clin Neurosci, 2018, 72(4): 195-211.

[104] Ng B, White CC, Klein HU, et al. An xQTL map integrates the genetic architecture of the human brain’s transcriptome and epigenome[J]. Nat Neurosci, 2017, 20(10): 1418-1426.

[105] Hannon E, Gorrie-Stone TJ, Smart MC, et al. Leveraging DNA-methylation quantitative-trait loci to characterize the relationship between methylomic variation, gene expression, and complex traits[J]. Am J Hum Genet, 2018, 103(5): 654-665.

[106] Li M, Li Y, Qin H, et al. Genome-wide DNA methylation analysis of peripheral blood cells derived from patients with first-episode schizophrenia in the Chinese Han population[J]. Mol Psychiatry, 2021, 26(8): 4475-4485.

[107] Shahryary Y, Hazarika RR, Johannes F. MethylStar: a fast and robust pre-processing pipeline for bulk or single-cell whole-genome bisulfite sequencing data[J]. BMC Genomics, 2020, 21(1): 479.

[108] Gershman A, Sauria MEG, Guitart X, et al. Epigenetic patterns in a complete human genome[J/OL]. Science, 2022, 376(6588): eabj5089[2025-05-12].

[109] Villicaña S, Bell JT. Genetic impacts on DNA methylation: research findings and future perspectives[J]. Genome Biol, 2021, 22(1): 127.

[110] Rizzardi LF, Hickey PF, Rodriguez DiBlasi V, et al. Neuronal brain-region-specific DNA methylation and chromatin accessibility are associated with neuropsychiatric trait heritability[J]. Nat Neurosci, 2019, 22(2): 307-316.

[111] Zhu Y, Gomez JA, Laufer BI, et al. Placental methylome reveals a 22q13.33 brain regulatory gene locus associated with autism[J]. Genome Biol, 2022, 23(1): 46.

[112] Mendizabal I, Berto S, Usui N, et al. Cell type-specific epigenetic links to schizophrenia risk in the brain[J]. Genome Biol, 2019, 20(1): 135.

[113] Tharakan R, Ubaida-Mohien C, Dunn C, et al. Whole-genome methylation analysis of aging human tissues identifies age-related changes in developmental and neurological pathways [J/OL]. Aging Cell, 2023, 22(7): e13847[2025-05-12].

[114] Breen C, Papale LA, Clark LR, et al. Whole genome methylation sequencing in blood identifies extensive differential DNA methylation in late-onset dementia due to Alzheimer’s disease[J]. Alzheimers Dement, 2024, 20(2): 1050-1062.

[115] Magwai T, Shangase KB, Oginga FO, et al. DNA methylation and schizophrenia: current literature and future perspective[J]. Cells, 2021, 10(11): 2890.

[116] Zhu JH, Bo HH, Liu BP, et al. The associations between DNA methylation and depression: a systematic review and meta-analysis[J]. J Affect Disord, 2023, 327: 439-450.

[117] Alshaya DS. Genetic and epigenetic factors associated with depression: an updated overview[J]. Saudi J Biol Sci, 2022, 29(8): 103311.

[118] Li L, Chen R, Zhang H, et al. The epigenetic modification of DNA methylation in neurological diseases[J]. Front Immunol, 2024, 15: 1401962.

[119] Behl T, Kyada A, Roopashree R, et al. Epigenetic biomarkers in Alzheimer’s disease: Diagnostic and prognostic relevance[J]. Ageing Res Rev, 2024, 102: 102556.

[120] Feng B, Zheng J, Cai Y, et al. An epigenetic manifestation of Alzheimer’s disease: DNA methylation[J]. Actas Esp Psiquiatr, 2024, 52(3): 365-374.

[121] De Plano LM, Saitta A, Oddo S, et al. Epigenetic changes in Alzheimer’s disease: DNA methylation and histone modification[J]. Cells, 2024, 13(8): 719.

[122] Song H, Chen J, Huang J, et al. Epigenetic modification in Parkinson’s disease[J]. Front Cell Dev Biol, 2023, 11: 1123621.

[123] Perzel Mandell KA, Eagles NJ, Wilton R, et al. Genome-wide sequencing-based identification of methylation quantitative trait loci and their role in schizophrenia risk[J]. Nat Commun, 2021, 12(1): 5251.

[124] Klengel T, Pape J, Binder EB, et al. The role of DNA methylation in stress-related psychiatric disorders[J]. Neuropharmacology, 2014, 80: 115-132.

[125] Kadayifci FZ, Zheng SS, Pan Y-X. Molecular mechanisms underlying the link between diet and DNA methylation[J]. Int J Mol Sci, 2018, 19(12): 4055.

[126] Rider CF, Carlsten C. Air pollution and DNA methylation: effects of exposure in humans[J]. Clin Epigenetics, 2019, 11(1): 131.

[127] Braun PR, Han SZ, Hing B, et al. Genome-wide DNA methylation comparison between live human brain and peripheral tissues within individuals[J]. Transl Psychiatry, 2019, 9(1): 47.

[128] Wei X, Zhang L, Zeng Y. DNA methylation in Alzheimer’s disease: in brain and peripheral blood[J]. Mech Ageing Dev, 2020, 191: 111319.

[129] Dolmetsch R, Geschwind DH. The human brain in a dish: the promise of iPSC-derived neurons[J]. Cell, 2011, 145(6): 831-834.

[130] Lee CT, Bendriem RM, Wu WW, et al. 3D brain Organoids derived from pluripotent stem cells: promising experimental models for brain development and neurodegenerative disorders[J]. J Biomed Sci, 2017, 24(1): 59.

[131] Brunner JW, Lammertse HCA, van Berkel AA, et al. Power and optimal study design in iPSC-based brain disease modelling[J]. Mol Psychiatry, 2023, 28(4): 1545-1556.

[132] Bartlett C, Bichindaritz I. Machine learning based classification of mental disorders from methylation data [C]//Innovation in Medicine and Healthcare Systems, and Multimedia. Singapore: Springer, 2019: 355-365.

[133] Gunasekara CJ, Hannon E, MacKay H, et al. A machine learning case-control classifier for schizophrenia based on DNA methylation in blood[J]. Transl Psychiatry, 2021, 11(1): 412.

[134] Cao Y, Bai Y, Yuan T, et al. Single-cell bisulfite-free 5mC and 5hmC sequencing with high sensitivity and scalability[J/OL]. Proc Natl Acad Sci USA, 2023, 120(49): e2310367120[2025-05-12].

[135] Zhang Q, Ma S, Liu ZZ, et al. Droplet-based bisulfite sequencing for high-throughput profiling of single-cell DNA methylomes[J]. Nat Commun, 2023, 14(1): 4672.

[136] Kurinomaru T, Kurita R. Bisulfite-free approaches for DNA methylation profiling[J]. Anal Methods, 2017, 9(10): 1537-1549.

[137] Simpson JT, Workman RE, Zuzarte PC, et al. Detecting DNA cytosine methylation using nanopore sequencing[J]. Nat Methods, 2017, 14(4): 407-410.

基金资助

the National Natural Science Foundation(82022024)

the Graduate Independent Innovation Project of Central South University(2022ZZTS0866)

RIGHTS & PERMISSIONS

©Journal of Central South University (Medical Science). All rights reserved.