Recent advances in cell-type deconvolution approaches are adding to our understanding of the biology underlying disease development and progression. DNA methylation (DNAm) can be used as a biomarker of cell types, and through deconvolution approaches, to infer underlying cell type proportions. Cell-type deconvolution algorithms have two main categories: reference-based and reference-free. Reference-based algorithms are supervised methods that determine the underlying composition of cell types within a sample by leveraging differentially methylated regions (DMRs) specific to cell type, identified from DNAm measures of purified cell populations. Reference-free algorithms are unsupervised methods for use when cell-type specific DMRs are not available, allowing scientists to estimate putative cellular proportions or control for potential confounding from cell type. Reference-based deconvolution is typically applied to blood samples and has potentiated our understanding of the relation between immune profiles and disease by allowing estimation of immune cell proportions from archival DNA. Bioinformatic analyses using DNAm to infer immune cell proportions, part of a new field known as Immunomethylomics, provides a new direction for consideration in epigenome wide association studies (EWAS).
Background: DNA methylation (DNAm) is an epigenetic regulator of gene expression programs that can be altered by environmental exposures, aging, and in pathogenesis. Traditional analyses that associate DNAm alterations with phenotypes suffer from multiple hypothesis testing and multi-collinearity due to the high-dimensional, continuous, interacting and non-linear nature of the data. Deep learning analyses have shown much promise to study disease heterogeneity. DNAm deep learning approaches have not yet been formalized into user-friendly frameworks for execution, training, and interpreting models. Here, we describe MethylNet, a DNAm deep learning method that can construct embeddings, make predictions, generate new data, and uncover unknown heterogeneity with minimal user supervision. Results: The results of our experiments indicate that MethylNet can study cellular differences, grasp higher order information of cancer sub-types, estimate age and capture factors associated with smoking in concordance with known differences. Conclusion: The ability of MethylNet to capture nonlinear interactions presents an opportunity for further study of unknown disease, cellular heterogeneity and aging processes.
OBJECTIVES To determine the relationship between frailty and overall and cardiovascular mortality. DESIGN Longitudinal mortality analysis. SETTING National Health and Nutrition Examination Survey (NHANES) 1999–2004. PARTICIPANTS Community-dwelling older adults aged 60 and older (N = 4,984; mean age 71.1 ± 0.19, 56% female). MEASUREMENTS We used data from 1999–2004 cross-sectional NHANES and mortality data from the National Death Index, updated through December 2011. An adapted version of Fried’s frailty criteria was used (low body mass index, slow walking speed, weakness, exhaustion, low physical activity). Frailty was defined as persons meeting 3 or more criteria, prefrailty as meeting 1 or 2 criteria, and robust (reference) as not meeting any criteria. The primary outcome was to evaluate the association between frailty and overall and cardiovascular mortality. Cox proportional hazard models were used to evaluate the association between risk of death and frailty category adjusted for age, sex, race, smoking, education, coronary artery disease, heart failure, nonskin cancer, diabetes, and arthritis. RESULTS Half (50.4%) of participants were classified as robust, 40.3% as prefrail, and 9.2% as frail. Fully adjusted models demonstrated that prefrail (hazard ratio (HR) = 1.64, 95% confidence interval (CI) = 1.45–1.85) and frail (HR = 2.79, 95% CI = 2.35–3.30) participants had a greater risk of death and of cardiovascular death (prefrail: HR = 1.84, 95% CI = 1.45–2.34; frail: HR = 3.39, 95% CI = 2.45–4.70). CONCLUSION Frailty and prefrailty are associated with increased risk of death. Demonstrating the association between prefrail status and mortality is the first step to identifying potential targets of intervention in future studies.
Objectives Sarcopenia is a gradual loss of muscle mass and strength that occurs with aging. This muscle deterioration is linked to increased morbidity, disability, and other adverse outcomes. Although reduced handgrip strength can be considered a marker of sarcopenia and other aging-related decline in the elderly, there is limited research on this physical health problem in at-risk groups with common biopsychosocial conditions such as depression. Our primary objective was to ascertain level of combined handgrip strength and its relationship with depression among adults aged 60 years and older. Design Unadjusted and adjusted linear regression models were conducted with a cross-sectional survey dataset. Setting Secondary dataset from the 2011–2014 National Health and Nutrition Examination Survey (NHANES). Participants Community-dwelling, non-institutionalized adults ≥ 60 years old (n=3,421). Measurements The predictor variables included a positive screen for clinically relevant depression (referent=PHQ-9 score <10). The criterion variable of combined handgrip strength (kg) was determined using a dynamometer. Results Mean age and BMI were 69.9 years (51.5% female) and 28.8 kg/m2, respectively. Mean combined handgrip strength in the overall cohort was 73.5 and 46.6 kg in males and females, respectively. Three hundred thirty-six (9.8%) reported symptoms of depression. In unadjusted and fully adjusted models, depression was significantly associated with reduced handgrip strength (B = −0.26±0.79 and B = −0.19±0.08, respectively; p<0.001). Conclusion Our findings demonstrate handgrip strength has a significant inverse association with depression. Future longitudinal studies should investigate the causal processes and potential moderators and mediators of the relationships between depression and reduced handgrip strength. This information may further encourage the use of depression and handgrip strength assessments and aid in the monitoring and implementation of health care services that address both physical and mental health limitations among older adult populations.
Objective: Body composition changes with aging can increase rates of obesity, frailty and impact function. Measuring adiposity using body fat (%BF) or central adiposity using waist circumference (WC) have greater diagnostic accuracy than traditional measures such as body mass index (BMI). Design: This is an observational study Setting: This study focused on older community-dwelling participants Participants: We identified individuals age ≥ 60 years old using the 1999–2004 cross-sectional National Health and Nutrition Survey (NHANES). Intervention: The primary analysis evaluated the association between frailty and %BF or WC. Frailty was the primary predictor (robust=referent) and %BF and WC were considered continuous outcomes. Multiple imputation analyses accounted for missing characteristics. Measurement: Dual energy x-ray absorptiometry was used to assess %BF and WC was objectively measured. Frailty was defined using an adapted version of Fried’s criteria that was self-reported: (low BMI<18.5kg/m2; slow walking speed [<0.8m/s]; weakness [unable to lift 10lbs]; exhaustion [difficulty walking between rooms on same floor] and low physical activity [compared to others]). Robust, pre-frail and frail persons met zero, 1 or 2, and ≥3 criteria, respectively. Results: Of the 4,984 participants, the mean age was 71.1±0.2 (SE) years and 56.5% were females. We classified 2,246 (50.4%), 2,195 (40.3%), and 541 (9.2%) individuals as robust, pre-frail and frail, respectively. Percent BF was 35.9±0.13, 38.3±0.20 and 40.0±0.46 in the robust, pre-frail and frail individuals, respectively. WC was 99.5±0.32 in the robust, 100.1±0.43 in pre-frail, 104.7±1.17 in frail individuals. Compared to robust individuals, only frail individuals had greater %BF on average (β=0.97±0.43,p=0.03); however, pre-frail and frail individuals had 2.18 and 4.80 greater WC, respectively (β=2.18±0.64,p=0.002, and β=4.80±1.1,p<0.001). Conclusion: Our results demonstrate that in older adults, frailty and pre-frailty are associated with a greater likelihood of high WC (as dichotomized) and a greater average WC (continuous).
Recent advances in deep learning, particularly unsupervised approaches, have shown promise for furthering our biological knowledge through their application to gene expression datasets, though applications to epigenomic data are lacking. Here, we employ an unsupervised deep learning framework with variational autoencoders (VAEs) to learn latent representations of the DNA methylation landscape from three independent breast tumor datasets. Through interrogation of methylation-based learned latent dimension activation values, we demonstrate the feasibility of VAEs to track representative differential methylation patterns among clinical subtypes of tumors. CpGs whose methylation was most correlated VAE latent dimension activation values were significantly enriched for CpG sparse regulatory regions of the genome including enhancer regions. In addition, through comparison with LASSO, we show the utility of the VAE approach for revealing novel information about CpG DNA methylation patterns in breast cancer.
Machine learning is a modern approach to problem-solving and task automation. In particular, machine learning is concerned with the development and applications of algorithms that PLOS COMPUTATIONAL BIOLOGY
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
334 Leonard St
Brooklyn, NY 11211
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.