DNA Methylation Data Research Articles

MethParquet: an R package for rapid and efficient DNA methylation association analysis adopting Apache Parquet.

Genome-wide DNA methylation (DNAm) profiling is indispensable for unveiling how DNAm regulates biological pathways and individual phenotypes. However, managing and analyzing extensive DNAm data generated from large cohort studies present computational obstacles. Apache Parquet is a data file format that allows for efficient data storage, retrieval, and manipulation, alleviating computational hurdles associated with conventional row-based formats. We here introduce MethParquet, the first R package leveraging the columnar Parquet format for efficient DNAm data analysis. It can be used for data extraction, methylation risk score calculation, epigenome-wide association analyses, and other standard post-quality control tasks. The package flexibly implements diverse regression models. Via a public methylation dataset, we show the efficiency of this package in reducing running time and RAM usage in large-scale EWAS. The MethParquet R package is publicly available on the GitHub repository https://github.com/ZWangTen/MethParquet. It includes a vignette and a toy dataset derived from a public resource.

MDB-71. IDENTIFICATION AND DESCRIPTION OF A NOVEL TYPE OF MEDULLOBLASTOMA

Abstract BACKGROUND Medulloblastoma (MB) is one of the most frequent high-grade brain tumor types of childhood and adolescence. Based on biology, histology, and its clinical course, MB is a heterogeneous disease with four different molecular types (WNT, SHH, Group 3, and Group 4), which are most reliably distinguishable by their global DNA methylation pattern. METHODS We analyzed DNA methylation (n=47), copy number variants, transcriptomic data (n=7) and histology (n=28) of a previously unrecognized type of MB. RESULTS As a result of integrating DNA methylation data of >2,600 MB and screening of >140,000 data sets uploaded to the DKFZ brain tumor classifier (www.molecularneuropathology.org), we identified a small group of MB that displayed a homogeneous DNA methylation pattern, which was clearly distinct from previously known MB types. Tumors within this group have also been recognized as a separate methylation class by latest versions of the classifier, which has provisionally been named medullomyoblastoma (MB_MYO). Transcriptomic data were similarly distinct from other MB types. Comparison of these data to various cell types of the developing hindbrain revealed transcriptional similarities to precursor cells of the rhombic lip with signatures of WNT signaling and myogenic differentiation. In line with the latter findings, 2/11 cases with DNA sequencing data harbored hotspot CTNNB1 mutations and 5/28 cases showed a myogenic differentiation based on histology. MYC amplifications were present in 11/47 cases, corresponding to a higher fraction of MYC-amplified tumors compared to all known MB types. Median age at diagnosis in this cohort was 16 years, and five-year overall survival was ~75 %. CONCLUSIONS In summary, we describe a novel type of MB identified by DNA methylation profiling that likely needs to be addressed separately during the retrospective analyses of MB patient cohorts, for the design of future clinical trials, and when evaluating targeted therapies.

ETMR-30. CLINICAL AND MOLECULAR CHARACTERIZATION OFDICER-MUTANT CENTRAL NERVOUS SYSTEM SARCOMA

Abstract BACKGROUND Central nervous system (CNS) sarcomas are rare mesenchymal non-meningothelial tumors accounting for 0.2% of brain tumors. Due to their rarity, there is a lack of substantive clinical, histological and molecular data. METHODS Patients diagnosed with primary CNS sarcomas were identified through several international collaborations. DNA methylation profiles were generated using 850K microarray assays, followed by unsupervised hierarchical clustering and t-SNE analyses. Additionally, targeted DNA sequencing was employed to characterize the somatic landscape. Clinical data was collected and summarized using descriptive statistics, and survival estimates were calculated using the Kaplan-Meier method. RESULTS We analyzed DNA methylation data of 65 tumor samples. Clustering analyses demonstrated DICER-1 CNS sarcomas segregate into two distinct epigenetic subgroups. Copy number analysis demonstrated frequent CNV alterations and chromothripsis in 35% of the samples. DICER-1 mutations were present in 85% of cases and associated with frequent additional alterations in TP53 (65%) and KRAS (51%). The median age at diagnosis was 10.4 years (r: 0.4 - 71 y) and 52% (n=34) of subjects were male. Tumors were frequently (91%) located in the supratentorial compartment, frontal lobe involvement was noted in 62% of cases. Patients were treated with multimodal approaches including surgery, ICE chemotherapy, and focal radiation. At median follow-up time of 27.5 months, the 2-year progression-free and overall survival was 50.8% (r: 0.31-0.67) and 53.5% (r: 0.34-0.69) respectively. CONCLUSIONS DICER-1 CNS Sarcoma is an aggressive entity that segregates into two epigenetic subgroups, clinical and genotype correlations remain to be elucidated. Current multimodal approaches are only effective in a subset of patients. The identification of MAPK pathway alterations in most cases should be explored as a therapeutic avenue.

Survival Rate and Chronic Diseases of TCGA Cancer and KoGES Normal Samples by Clustering for DNA Methylation.

Insights from public DNA methylation data derived from cancer or normal tissues from cancer patients or healthy people can be obtained by machine learning. The goal is to determine methylation patterns that could be useful for predicting the prognosis for cancer patients and correcting lifestyles for healthy people. DNA methylation data were obtained from the DNA of 446 healthy participants from the Korean Genome Epidemiology Study (KoGES) and from the DNA of normal tissues or from cancer tissues of 11 types of carcinomas from The Cancer Genome Atlas (TCGA) database. To correct for the batch effect, R's ComBat function was used. Using the K-mean clustering (k = 3), the survival rates of the cancer patients and the incidence of chronic diseases were compared between the three clusters for TCGA and KoGES, respectively. Based on the public DNA methylation and clinical data of healthy participants and cancer patients, I present an analysis pipeline that integrates and clusters the methylation data from the two groups. As a result of clustering, CpG sites from gene or genomic regions, such as AFAP1, NINJ2, and HOOK2 genes, that correlated with survival rate and chronic disease are presented.

MethylGenotyper: Accurate Estimation of SNP Genotypes and Genetic Relatedness from DNA Methylation Data.

Epigenome-wide association studies (EWAS) are susceptible to widespread confounding caused by population structure and genetic relatedness. Nevertheless, kinship estimation is challenging in EWAS without genotyping data. Here, we proposed MethylGenotyper, a method that for the first time enables accurate genotyping at thousands of single nucleotide polymorphisms (SNPs) directly from commercial DNA methylation microarrays. We modeled the intensities of methylation probes near SNPs with a mixture of three beta distributions corresponding to different genotypes and estimated parameters with an expectation-maximization algorithm. We conducted extensive simulations to demonstrate the performance of the method. When applying MethylGenotyper to the Infinium EPIC array data of 4662 Chinese samples, we obtained genotypes at 4319 SNPs with a concordance rate of 98.26%, enabling the identification of 255 pairs of close relatedness. Furthermore, we showed that MethylGenotyper allows for the estimation of both population structure and cryptic relatedness among 702 Australians of diverse ancestry. We also implemented MethylGenotyper in a publicly available R package (https://github.com/Yi-Jiang/MethylGenotyper) to facilitate future large-scale EWAS.

Exploration of mitochondrial autophagy related genes in the diagnosis model construction and molecular marker mining of Alzheimer's disease based on multi-omics integration.

Key features of Alzheimer's disease include neuronal loss, accumulation of beta-amyloid plaques, and formation of neurofibrillary tangles. These changes are due in part to abnormal protein metabolism, particularly the accumulation of amyloid beta. Mitochondria are the energy production centers within cells and are also the main source of oxidative stress. In AD, mitochondrial function is impaired, leading to increased oxidative stress and the production of more reactive oxidative substances, further damaging cells. Mitophagy is an important mechanism for maintaining mitochondrial health, helping to clear damaged mitochondria, prevent the spread of oxidative stress, and reduce abnormal protein aggregation. To this end, this article conducts an integrated analysis based on DNA methylation and transcriptome data of AD. After taking the intersection of the genes where the differential methylation sites are located and the differential genes, machine learning methods were used to build an AD diagnostic model. This article screened five diagnostic genes ATG12, CSNK2A2, CSNK2B, MFN1 and PGAM5 and conducted experimental verification. The diagnostic genes discovered and the diagnostic model constructed in this article can provide reference for the development of clinical diagnostic models for AD.

Precise Molecular Subtyping Reveals Heterogeneity of Lung Adenocarcinoma Based on DNA Methylation

Background: Due to the high heterogeneity of lung adenocarcinoma (LUAD), which restricts the effectiveness of therapy, precise molecular subgrouping of LUAD is of great significance. Clinical research has demonstrated the significant potential of DNA methylation as a classification indicator for human malignancies. Methods: WGML framework (which was developed based on weighted gene correlation network analysis (WGCNA), Gene Ontology (GO), and machine learning) was developed to precisely subgroup molecular subtypes of LUAD. This framework included two parts: the WG algorithm and the machine learning part. The WG algorithm part was an original algorithm used to obtain a crucial module, which was characterized by weighted correlation network analysis, functional annotation, and mathematical algorithms. The machine learning part utilized the Boruta algorithm, random forest algorithm, and Gradient Boosting Regression Tree algorithm to select feature genes. Then, based on the results of the WGML framework, subtypes were computed by the hierarchical clustering algorithm. A series of analyses, including dimensionality reduction methods, survival analysis, clinical stage analysis, immune infiltration analysis, tumor environment analysis, immune checkpoints analysis, TIDE analysis, CYT analysis, somatic mutation analysis, and drug sensitivity analysis, were utilized to demonstrate the effectiveness of subgrouping. GEO datasets were used to externally validate the results. Meanwhile, another subgrouping method of LUAD from another study was employed to compare with the WGML framework. Result: By importing DNA methylation data into the WGML framework, nine genes were obtained to further subgroup LUAD. Three subtypes, the Carcinogenesis subtype, Immune-infiltration subtype, and Chemoresistance subtype, were identified. The dimensionality reduction method exhibited great distinctness between subtypes. A series of analyses were employed to exhibit the difference among the three subtypes and to demonstrate the accuracy of the definition of subtypes. Besides, the WGML framework was compared with a LUAD subgrouping method from another research, which demonstrated that WGML had better efficiency for subgrouping LUAD. Conclusion: This study provides a novel LUAD subgrouping framework named WGML for the accurate subgrouping of lung adenocarcinoma. result: By importing DNA methylation data into WGML framework, nine genes were obtained to further subgroup LUAD. Three subtypes, Carcinogenesis subtype, Immune-infiltration subtype, and Chemoresistance subtype were identified. Dimensionality reduction method exhibited great distinctness between subtypes. A series analyses were employed to exhibit the difference of three subtypes and to demonstrate accuracy of definition of subtypes. Besides, WGML framework was compared with a LUAD subgrouping method from another research, which demonstrated WGML had better efficiency for subgrouping LUAD.

Smoking-informed methylation and expression QTLs in human brain and colocalization with smoking-associated genetic loci

Smoking is a leading cause of preventable morbidity and mortality. Smoking is heritable, and genome-wide association studies (GWASs) of smoking behaviors have identified hundreds of significant loci. Most GWAS-identified variants are noncoding with unknown neurobiological effects. We used genome-wide genotype, DNA methylation, and RNA sequencing data in postmortem human nucleus accumbens (NAc) to identify cis-methylation/expression quantitative trait loci (meQTLs/eQTLs), investigate variant-by-cigarette smoking interactions across the genome, and overlay QTL evidence at smoking GWAS-identified loci to evaluate their regulatory potential. Active smokers (N = 52) and nonsmokers (N = 171) were defined based on cotinine biomarker levels and next-of-kin reporting. We simultaneously tested variant and variant-by-smoking interaction effects on methylation and expression, separately, adjusting for biological and technical covariates and correcting for multiple testing using a two-stage procedure. We found >2 million significant meQTL variants (padj < 0.05) corresponding to 41,695 unique CpGs. Results were largely driven by main effects, and five meQTLs, mapping to NUDT12, FAM53B, RNF39, and ADRA1B, showed a significant interaction with smoking. We found 57,683 significant eQTL variants for 958 unique eGenes (padj < 0.05) and no smoking interactions. Colocalization analyses identified loci with smoking-associated GWAS variants that overlapped meQTLs/eQTLs, suggesting that these heritable factors may influence smoking behaviors through functional effects on methylation/expression. One locus containing MUSTN1 and ITIH4 colocalized across all data types (GWAS, meQTL, and eQTL). In this first genome-wide meQTL map in the human NAc, the enriched overlap with smoking GWAS-identified genetic loci provides evidence that gene regulation in the brain helps explain the neurobiology of smoking behaviors.

Causal impact of DNA methylation on refracture in elderly individuals with osteoporosis – a prospective cohort study

BackgroundOsteoporotic vertebral compression fractures (OVCF) in the elderly increase refracture risk post-surgery, leading to higher mortality rates. Genome-wide association studies (GWAS) have identified susceptibility genes for osteoporosis, but the phenotypic variance explained by these genes has been limited, indicating the need to explore additional causal factors. Epigenetic modifications, such as DNA methylation, may influence osteoporosis and refracture risk. However, prospective cohorts for assessing epigenetic alterations in Chinese elderly patients are lacking. Here, we propose to conduct a prospective cohort study to investigate the causal network of DNA polymorphisms, DNA methylation, and environmental factors on the development of osteoporosis and the risk of refracture.MethodsWe will collect vertebral and peripheral blood from 500 elderly OVCF patients undergoing surgery, extract DNA, and generate whole genome genotype data and DNA methylation data. Observation indicators will be collected and combined with one-year follow-up data. A healthy control group will be selected from a natural population cohort. Epigenome-wide association studies (EWAS) of osteoporosis and bone mineral density will be conducted. Differential methylation analysis will compare candidate gene methylation patterns in patients with and without refracture. Multi-omics prediction models using genetic variants and DNA methylation sites will be built to predict OVCF risk.DiscussionThis study will be the first large-scale population-based study of osteoporosis and bone mineral density phenotypes based on genome-wide data, multi-time point methylation data, and phenotype data. By analyzing methylation changes related to osteoporosis and bone mineral density in OVCF patients, the study will explore the feasibility of DNA methylation in evaluating postoperative osteoporosis intervention effects. The findings may identify new molecular markers for effective anti-osteoporosis treatment and inform individualized prevention and treatment strategies.Trial registrationchictr.org.cn ChiCTR2200065316, 02/11/2022.

MCDHGN: heterogeneous network-based cancer driver gene prediction and interpretability analysis.

Accurately predicting the driver genes of cancer is of great significance for carcinogenesis progress research and cancer treatment. In recent years, more and more deep-learning-based methods have been used for predicting cancer driver genes. However, deep-learning algorithms often have black box properties and cannot interpret the output results. Here, we propose a novel cancer driver gene mining method based on heterogeneous network meta-paths (MCDHGN), which uses meta-path aggregation to enhance the interpretability of predictions. MCDHGN constructs a heterogeneous network by using several types of multi-omics data that are biologically linked to genes. And the differential probabilities of SNV, DNA methylation, and gene expression data between cancerous tissues and normal tissues are extracted as initial features of genes. Nine meta-paths are manually selected, and the representation vectors obtained by aggregating information within and across meta-path nodes are used as new features for subsequent classification and prediction tasks. By comparing with eight homogeneous and heterogeneous network models on two pan-cancer datasets, MCDHGN has better performance on AUC and AUPR values. Additionally, MCDHGN provides interpretability of predicted cancer driver genes through the varying weights of biologically meaningful meta-paths. https://github.com/1160300611/MCDHGN.

A random forest model to predict the efficacy of anti-PD-1/PD-L1 monoclonal antibody in lung adenocarcinoma.

e20558 Background: Not all patients with lung adenocarcinoma (LUAD) could benefit from anti-PD-1/PD-L1 monoclonal antibody therapy, and existing biomarkers cannot explain all clinical benefits. Therefore, an effective model to predict the efficacy of anti-PD-1/PD-L1 monoclonal antibody is necessary. Methods: DNA methylation sequencing was used in LUAD patients’ FFPE samples, who received anti-PD-1/PD-L1 monoclonal antibody. The DNA methylation level of each gene was transformed into binary format according to its median value. The progression-free survival (PFS) was defined as the duration from the beginning of anti-PD-1/PD-L1 monoclonal antibody therapy to disease progression or death. Patients were randomly divided into training cohort (70%) and test cohort (30%). In the training cohort, uni-Cox regression was used to estimate the value in predicting PFS, and LASSO-Cox regression was performed to screened genes to build random forest model. The accuracy of this random forest model was validated in the test cohort. Finally, DNA methylation data were used to calculated the risk score of patients in the TCGA-LUAD cohort, and GSEA and GSVA were employed in RNA data to explore the mechanism and explain the efficacy of the prediction model. Results: In this study, 40 patients with LUAD were enrolled, and the median PFS of them was 7.57 months. A total of 29 genes with P &lt; 0.05 and C-index &gt; 0.7 were selected to perform LASSO-Cox regression, and 13 genes ( BTBD7, CAD, FBXL20, FKBP7, HBP1, IGFBP3, RAB5B, RASGRP1, RNF8, SOHLH1, TMEM223, UHRF1 and WWC2) were used to build prediction model. The C-index of risk score from this model was 0.96 and 0.93 in the training and test cohort, respectively. Patients were divided into the high and low risk groups according to the median risk score, and the median PFS of the high risk group was shorter than that of the low risk group (training cohort: 3.05 vs. 10.89 months, P &lt; 0.01, HR = 2.51, 95% CI: 1.50-4.22; test cohort: 1.64 vs. 11.31 months, P = 0.03, HR = 1.80, 95% CI: 1.04-3.14). In the low and high risk group of the TCGA-LUAD cohort, there were 92.2% (388/421) and 0.5% (2/421) significant enriched ( P &lt; 0.05 and FDR &lt; 0.25) immunity-related pathways from results of GSEA, respectively, meanwhile 94.3% (397/421) and 1.9% (8/421) significant enriched immunity-related pathways from results of GSVA, respectively. Therefore, the lower risk score of this random forest model accompanied with the higher activation of immunity-related pathways which might help anti-PD-1/PD-L1 monoclonal antibody to express better efficacy. Conclusions: A random forest model based on DNA methylation data was constructed and showed a potential value in predicting the PFS of anti-PD-1/PD-L1 monoclonal antibody therapy in LUADs, and the risk score of this model is negatively correlated to the activation of immunity-related pathways, which could help to identify LUAD patients that would benefit from immunotherapy.

Novel epigenomic liquid biopsy assay to predict estrogen receptor (ER) status and to infer ER pathway activation in breast cancer.

1066 Background: Endocrine therapy (ET) is the cornerstone of treatment for ER-positive (ER+) breast cancer (BC). However, BC may lose ER expression over the disease course, and ER+ tumors may become resistant to ET regardless of ER expression. Blood-based biomarkers to infer ER expression and to predict ET sensitivity are lacking. Methods: We identified 44 patients (pts) with metastatic BC seen at Dana-Farber Cancer Institute from 2012-2023, who had blood samples drawn in Streck tubes within 6 weeks of tumor biopsy with ER scored via immunohistochemistry (IHC). We applied a novel, multi-analyte, liquid biopsy assay to capture genome-wide epigenomic signals, mapping enhancers, promoters, and DNA methylation data from 1mL of plasma. First, we applied a previously developed, IHC-benchmarked, regularized logistic regression model to infer ER status from plasma-based epigenomic profiles (Parsons et al., SABCS 2023). Second, we applied a novel ER pathway activity score involving promoters and enhancers of certain genes previously associated with ER activity (Guan et al., Cell 2019). The ER activity score was corrected for the circulating tumor DNA (ctDNA) fraction as assessed by low pass whole genome sequencing (ichorCNA algorithm). We evaluated the association between the ER activity score and the ER status by IHC using the Wilcoxon rank-sum test. ER expression ≥1% was considered positive. Results: The ER classifier and the ER activity score have been applied to 19 of 44 samples (43%) all of which had detectable ctDNA. Eleven out of 19 (58%) pts had ER+ BC by IHC (ER IHC range 80-95%). Overall, 18/19 (95%) were correctly classified as ER+ or ER- by the ER classifier (AUC=1). Three of 19 (16%) samples were collected from pts with ER+ BC at primary diagnosis and ER- BC at time of tissue sampling. For them, the ER classifier correctly predicted ER- status. We observed a significantly higher ER activity score in samples from pts with ER+ BC than pts with ER- BC as measured by IHC (p= 0.0012). For pts with ER+ BC, we observed an association between lower ER activity scores and a higher number of lines of ET prior to blood sampling, with lower scores for pts who received &gt;4 prior lines. We will present full results from this cohort and an expanded cohort and correlate the ER score with response to ET. Conclusions: We applied a novel blood-based epigenomic classifier able to infer ER status and a novel epigenomic activity score to predict ER signaling from a minimal plasma volume. We plan to validate our findings on a larger cohort and correlate with response to ET. If confirmed, this assay may represent a valuable, minimally invasive tool to help guide treatment selection for patients with ER+ BC and predict response to novel endocrine agents.

High-sensitive multi-cancers early detection for 8 cancer types by cell-free DNA targeted methylation sequencing: A retrospective cohort study.

3043 Background: Early screening effectively reduces mortality associated with cancer. Currently, most cancers lack effective screening paradigms. We developed blood-based multi-cancers early detection (MCED) and cancer signal origin (CSO) approaches in 8 types of cancers including esophageal, stomach, colorectal, pancreatic, liver, lung, breast and ovarian cancer. Methods: DNA methylation data generated based on GM-seq by public and internal whole genome methylation data. Based on these data, a panel of 155,362 CpG sites spanning the 2.0M genome was constructed and validated. We enrolled a case-control cohort of 746 participants (522 cancers, 224 non-cancers) for MCED model development. cell-free DNA (cfDNA) was isolated from 10 ml peripheral blood from each participant. A targeted methylation sequencing of cfDNA in plasma baseline was established using 224 non-cancers blood samples for 8 types of cancers. Finally, we developed a MCED model named AMBER for distinguishing cancer from non-cancer individuals. Results: DNA methylation data were from 8 types of cancers (cancer tissues: n=812, adjacent/normal tissues: n=416, cancer peripheral bloods: n=624, normal peripheral bloods: n=503) in internal and Infinium Human Methylation 450K array (cancer tissues: n=5000+, adjacent/normal tissues: n=1000+) from public data. 224 non-cancer individuals (healthy: n=196, cancer benign: n=28, male: female, 54.3%: 45.7%) were collected as control, while 522 cancer patients (male: female, 50.0%: 50.0%, stage I 26.9%, II 21.1%) from esophageal (n=56, stage I: n=10), stomach (n=47, stage I: n=11), colorectal (n=72, stage I: n=16), pancreatic (n=64, stage I: n=16), liver (n=69, stage I: n=22), lung (n=103, stage I: n=46), breast (n=41) and ovarian cancer (n=70, stage I: n=17). AMBER showed 99.1% specificity and total sensitivity 77.2% (95% confidence interval (CI): 73.4% to 80.7%). Of them, 56.8% (95% CI: 48.2% to 65.2%) of all stage I cases were detected, 70.6% (95% CI: 61.2% to 79.0%) for stage II, 84.5% (95% CI: 77.6% to 89.9%) for stage III, and 96.7% (95% CI: 91.8% to 99.1%) for stage IV. Strikingly, for stage I cancers, the sensitivity of esophageal, stomach, colorectal, pancreatic, liver, lung and ovarian cancer were 80.0%, 63.6%, 43.8%, 62.5%, 86.4%, 30.4% and 82.4% respectively. The sensitivity of esophageal, stomach, colorectal, pancreatic, liver, lung, breast and ovarian cancer reached 91.1%, 76.6%, 75.0%, 78.1%, 94.2%, 55.3%, 65.9% and 90.0%. For CSO model, the accuracy of TPO1 and TPO2 was 78.9% (95% CI: 74.6% to 82.8%) and 89.1% (95% CI: 85.6% to 92.0%). Conclusions: Our MCED tests demonstrated superior performance in detecting 8 types of cancers, especially early cancer screening, by utilizing cfDNA methylation information. It suggests the model can complement lack of multi-cancers early detection.

Integrative analyses of multi-omics data constructing tumor microenvironment and immune-related molecular prognosis model in human colorectal cancer

The increasing prevalence and incidence of colorectal cancer (CRC), particularly in young adults, underscore the imperative to comprehend its fundamental mechanisms, discover novel diagnostic and prognostic markers, and enhance therapeutic strategies. Here, we integrated multi-omics data, including gene expression, somatic mutation data and DNA methylation data, to unravel the intricacies of tumor microenvironment (TME) in CRC and search for novel prognostic markers. By calculating the immune score for each patient from the expression profile, we delineated the differential immune cell fraction, constructed an immune-related multi-omics atlas, and identified molecular characteristics. The entire colorectal dataset (n = 343) was randomly divided into training (n = 249) and testing datasets (n = 94). We screened 144 immune-related genes, 6 mutant genes, and 38 methylation probes associated with overall survival (OS). These makers were then incorporated into a 10-gene prognostic model using Lasso and Cox regression in the training dataset, and the model's performance was evaluated in an independent validation dataset. The model exhibited satisfactory results (average concordance index [C-index] = 0.77), with the average 1-year, 3-year, and 5-year AUCs being 0.79, 0.76, and 0.76 in the training dataset and 0.74, 0.80, and 0.90 in the testing dataset. Furthermore, the prognostic model demonstrated applicability in guiding chemotherapy for CRC patients and exhibited a degree of pan-cancer utility in risk stratification. In conclusion, our integrated analysis of multi-omics data revealed immune-related genetic and epigenetic characteristics of the TME. We propose an integrative prognostic model that can stratify risk and guide chemotherapy for CRC patients. The generalizability of the model in risk stratification across different cancer types was validated in Pan-Cancer cohort.

Chimeric antigen receptor T-cell therapy in B-cell malignancies and DNA-methylation–based biological aging.

2562 Background: Chimeric antigen receptor T-cell (CAR T-cell) therapy, a type of cancer immunotherapy which leverages genetically altered T cells to target cancer cells, has shown to be an effective option for patients with acute lymphoblastic leukemia (ALL) and non-Hodgkin lymphoma (NHL). Despite its growing use, the effect of CAR T-cell therapy on the aging process remains limited and findings may have implications for disease relapse and age-related diseases among survivors. The purpose of this study was to investigate the association between CAR T-cell therapy and markers of biological aging. Methods: DNA methylation and clinical data from a publicly available dataset (GSE179414) were used in this analysis. Transduced and untransduced T cells from patients with ALL (n=77) and NHL (n=37) underwent DNA methylation profiling using the Illumina Infinium MethylationEPIC BeadChip. Six DNA methylation-based biological age metrics were calculated: intrinsic epigenetic age acceleration (IEAA), extrinsic epigenetic age acceleration (EEAA), PhenoAge acceleration (PhenoAA), GrimAge acceleration (GrimAA), telomere length attrition (TLA), and DunedinPACE (Pace of Aging Calculated from the Epigenome). Multiple linear regression analyses were performed to examine the association between CAR T-cell therapy and the aging metrics by B-cell malignancy. Interaction and stratified analyses by chronological age was additionally performed. Results: DunedinPACE ( P=0.002) was positively associated with CAR T-cell therapy among patients with ALL with an approximately 1-month higher pace of aging in transduced T cells compared to untransduced T cells after adjusting for chronological age and sex. Moreover, chronological age significantly ( P=0.016) modified the association between CAR T-cell therapy and TLA among ALL patients. Specifically, transduced T cells exhibited a 0.46 higher ( P=0.053) and a 0.39 lower ( P=0.142) rate of TLA among patients 16 years of age or older and those younger than 16 years of age, respectively. Conclusions: We observed CAR T-cell therapy was associated with greater biological aging as estimated from DNA methylation among ALL patients and these associations are modified by chronological age. These findings suggest CAR T-cell therapy may be associated with age-related changes to the epigenome in ALL patients and strategies to limit or reverse this effect may have implications for the aging process in cancer survivors. [Table: see text]

DNA methylation-based biological age and time to relapse in triple negative breast cancer.

e13126 Background: Triple negative breast cancer (TNBC) represents 15% of all breast cancer cases and is one of the most aggressive forms of breast cancer with a poorer prognosis. Due to the lack of receptors, TNBC is not amendable to hormonal or anti-HER2 therapies and current gene expression arrays have limited clinical utility in TNBC. The identification of biomarkers associated with TNBC may aid in elucidating the underlying biological mechanisms of this disease and may also have implications for risk stratification, diagnosis, and prognosis of TNBC. The purpose of this study was to investigate the association between markers of biological aging and time to relapse in TNBC. Methods: DNA methylation and clinical data from a publicly available dataset (GSE141441) were used in this analysis. Breast tissue specimens in female TNBC patients who received (n=56) and who did not receive (n=110) chemotherapy underwent DNA methylation profiling using the Illumina Infinium HumanMethylation450K BeadChip. Five DNA methylation-based biological age metrics were calculated: intrinsic epigenetic age acceleration (IEAA), extrinsic epigenetic age acceleration (EEAA), PhenoAge acceleration (PhenoAA), GrimAge acceleration (GrimAA), and DunedinPACE (Pace of Aging Calculated from the Epigenome). Multiple linear and quantile regression analyses were performed to examine the association between the aging metrics and time to TNBC relapse. Interaction and stratified analyses by Ki-67 was additionally performed. Results: After adjusting for chronological age, GrimAA ( P=0.024) and DunedinPACE ( P=0.019) were negatively associated with time to relapse among patients who did not receive chemotherapy. Specifically, a 1-year increase in GrimAA and 1 unit increase in the pace of aging were associated with a 1.77-month and a 53.62-month shorter relapse of TNBC, respectively. Additionally, Ki-67 significantly modified the association between IEAA ( P=0.023) and DunedinPACE ( P=0.035) among patients who received chemotherapy with a 2.75- and a 42.17-month shorter relapse of TNBC among those with a Ki-67 index greater than 50% per unit increase in IEAA and DunedinPACE, respectively. Conclusions: We observed DNA methylation markers of biological age were associated with time to relapse in TNBC patients and these associations were modified by Ki-67. These biomarkers may have prognostic utility in TNBC relapse and potentially, may aid in oncological precision medicine efforts for personalized treatment regimens and surveillance. [Table: see text]

Prognosis and diagnosis of prostate cancer based on hypergraph regularization sparse least partial squares regression algorithm.

Prostate cancer (PCa) is a malignant tumor of the male reproductive system, and its incidence has increased significantly in recent years. This study aimed to further identify candidate biomarkers with prognostic and diagnostic significance by integrating gene expression and DNA methylation data from PCa patients through association analysis. To this end, this paper proposes a sparse partial least squares regression algorithm based on hypergraph regularization (HR-SPLS) by integrating and clustering two kinds of data. Next, module 2, with the most significant weight, was selected for further analysis according to the weight of each module related to DNA methylation and mRNAs. Based on the DNA methylation sites in module 2, this paper uses multiple machine learning methods to construct a PCa diagnosis-related model of 10-DNA methylation sites. The results of Receiver Operating Characteristic (ROC) analysis showed that the DNA methylation-related diagnostic model we constructed could diagnose PCa patients with high accuracy. Subsequently, based on the mRNAs in module 2, we constructed a prognostic model for 7-mRNAs (MYH11, ACTG2, DDR2, CDC42EP3, MARCKSL1, LMOD1, and MYLK) using multivariate Cox regression analysis. The prognostic model could predict the disease free survival of PCa patients with moderate to high accuracy (area under the curve (AUC) =0.761). In addition, Gene Set EnrichmentAnalysis (GSEA) and immune analysis indicated that the prognosis of patients in the risk group might be related to immune cell infiltration. Our findings may provide new methods and insights for identifying disease-related biomarkers by integrating DNA methylation and gene expression data.

Multi-omics in nasal epithelium reveals three axes of dysregulation for asthma risk in the African Diaspora populations

Asthma has striking disparities across ancestral groups, but the molecular underpinning of these differences is poorly understood and minimally studied. A goal of the Consortium on Asthma among African-ancestry Populations in the Americas (CAAPA) is to understand multi-omic signatures of asthma focusing on populations of African ancestry. RNASeq and DNA methylation data are generated from nasal epithelium including cases (current asthma, N = 253) and controls (never-asthma, N = 283) from 7 different geographic sites to identify differentially expressed genes (DEGs) and gene networks. We identify 389 DEGs; the top DEG, FN1, was downregulated in cases (q = 3.26 × 10−9) and encodes fibronectin which plays a role in wound healing. The top three gene expression modules implicate networks related to immune response (CEACAM5; p = 9.62 × 10−16 and CPA3; p = 2.39 × 10−14) and wound healing (FN1; p = 7.63 × 10−9). Multi-omic analysis identifies FKBP5, a co-chaperone of glucocorticoid receptor signaling known to be involved in drug response in asthma, where the association between nasal epithelium gene expression is likely regulated by methylation and is associated with increased use of inhaled corticosteroids. This work reveals molecular dysregulation on three axes – increased Th2 inflammation, decreased capacity for wound healing, and impaired drug response – that may play a critical role in asthma within the African Diaspora.

Self-rated health, epigenetic ageing, and long-term mortality in older Australians.

Self-rated health (SRH) is a subjective indicator of overall health based on a single questionnaire item. Previous evidence found that it is a strong predictor of mortality, although the underlying mechanism is poorly understood. Epigenetic age is an objective, emerging biomarker of health, estimated using DNA methylation data at hundreds of sites across the genome. This study aimed to assess the overlap and interaction between SRH and epigenetic ageing in predicting mortality risk. We used DNA methylation data from 1059 participants in the Melbourne Collaborative Cohort Study (mean age: 69years) to calculate three age-adjusted measures of epigenetic ageing: GrimAge, PhenoAge, and DunedinPACE. SRH was assessed using a five-category questionnaire item ("excellent, very good, good, fair, poor"). Cox models were used to assess the associations of SRH, epigenetic ageing, and their interaction, with all-cause mortality over up to 17years of follow-up (Ndeaths = 345). The association of SRH with mortality per category increase was HR = 1.29; 95%CI: 1.14-1.46. The association was slightly attenuated after adjusting for all three epigenetic ageing measures (HR = 1.25, 95%CI: 1.10-1.41). A strong gradient was observed in the association of GrimAge (Pinteraction = 0.006) and DunedinPACE (Pinteraction = 0.002) with mortality across worsening SRH strata. For example, the association between DunedinPACE and mortality in participants with "excellent" SRH was HR = 1.02, 95%CI: 0.73-1.43 and for "fair/poor" HR = 1.72, 95%CI: 1.35-2.20. SRH and epigenetic ageing were synergistic risk factors of mortality in our study. These findings suggest that consideration of subjective and objective factors may improve general health assessment, which has implications for the ongoing development of molecular markers of ageing.

Healthy Japanese dietary pattern is associated with slower biological aging in older men: WASEDA’S health study

Aging is the greatest risk factor for numerous diseases and mortality, and establishing geroprotective interventions targeting aging is required. Previous studies have suggested that healthy dietary patterns, such as the Mediterranean diet, are associated with delayed biological aging; however, these associations depend on nationality and sex. Therefore, this study aimed to investigate the relationship between dietary patterns identified through principal component analysis and biological aging in older men of Japan, one of the countries with the longest life expectancies. Principal component analysis identified two dietary patterns: a healthy Japanese dietary pattern and a Western-style dietary pattern. Eight epigenetic clocks, some of the most accurate aging biomarkers, were identified using DNA methylation data from whole-blood samples. Correlation analyses revealed that healthy Japanese dietary patterns were significantly negatively or positively correlated with multiple epigenetic age accelerations (AgeAccel), including AgeAccelGrim, FitAgeAccel, and age-adjusted DNAm-based telomere length (DNAmTLAdjAge). Conversely, the Western-style dietary pattern was observed not to correlate significantly with any of the examined AgeAccels or age-adjusted values. After adjusting for covariates, the healthy Japanese dietary pattern remained significantly positively correlated with DNAmTLAdjAge. Regression analysis showed that healthy Japanese dietary pattern contributed less to epigenetic age acceleration than smoking status. These findings suggest that a Western-style dietary pattern may not be associated with biological aging, whereas a healthy Japanese dietary pattern is associated with delayed biological aging in older Japanese men. Our findings provide evidence that healthy dietary patterns may have mild beneficial effects on delayed biological aging in older Japanese men.