Predicting genome-wide DNA methylation using methylation marks, genomic position, and DNA regulatory elements

Epigenetic Landscapes of Pain: DNA Methylation Dynamics in Chronic Pain

Chronic pain is a prevalent condition with a multifaceted pathogenesis, where epigenetic modifications, particularly DNA methylation, might play an important role. This review delves into the intricate mechanisms by which DNA methylation and demethylation regulate genes associated with nociception and pain perception in nociceptive pathways. We explore the dynamic nature of these epigenetic processes, mediated by DNA methyltransferases (DNMTs) and ten-eleven translocation (TET) enzymes, which modulate the expression of pro- and anti-nociceptive genes. Aberrant DNA methylation profiles have been observed in patients with various chronic pain syndromes, correlating with hypersensitivity to painful stimuli, neuronal hyperexcitability, and inflammatory responses. Genome-wide analyses shed light on differentially methylated regions and genes that could serve as potential biomarkers for chronic pain in the epigenetic landscape. The transition from acute to chronic pain is marked by rapid DNA methylation reprogramming, suggesting its potential role in pain chronicity. This review highlights the importance of understanding the temporal dynamics of DNA methylation during this transition to develop targeted therapeutic interventions. Reversing pathological DNA methylation patterns through epigenetic therapies emerges as a promising strategy for pain management.

Differential methylation region detection via an array-adaptive normalized kernel-weighted model

A differentially methylated region (DMR) is a genomic region that has significantly different methylation patterns between biological conditions. Identifying DMRs between different biological conditions is critical for developing disease biomarkers. Although methods for detecting DMRs in microarray data have been introduced, developing methods with high precision, recall, and accuracy in determining the true length of DMRs remains a challenge. In this study, we propose a normalized kernel-weighted model to account for similar methylation profiles using the relative probe distance from "nearby" CpG sites. We also extend this model by proposing an array-adaptive version in attempt to account for the differences in probe spacing between Illumina's Infinium 450K and EPIC bead array respectively. We also study the asymptotic results of our proposed statistic. We compare our approach with a popular DMR detection method via simulation studies under large and small treatment effect settings. We also discuss the susceptibility of our method in detecting the true length of the DMRs under these two settings. Lastly, we demonstrate the biological usefulness of our method when combined with pathway analysis methods on oral cancer data. We have created an R package called idDMR, downloadable from GitHub repository with link: https://github.com/DanielAlhassan/idDMR, that allows for the convenient implementation of our array-adaptive DMR method.

Integration of autoencoder and graph convolutional network for predicting breast cancer drug response

Background and objectives: Breast cancer is the most prevalent type of cancer among women. The effectiveness of anticancer pharmacological therapy may get adversely affected by tumor heterogeneity that includes genetic and transcriptomic features. This leads to clinical variability in patient response to therapeutic drugs. Anticancer drug design and cancer understanding require precise identification of cancer drug responses. The performance of drug response prediction models can be improved by integrating multi-omics data and drug structure data. Methods: In this paper, we propose an Autoencoder (AE) and Graph Convolutional Network (AGCN) for drug response prediction, which integrates multi-omics data and drug structure data. Specifically, we first converted the high dimensional representation of each omic data to a lower dimensional representation using an AE for each omic data set. Subsequently, these individual features are combined with drug structure data obtained using a Graph Convolutional Network and given to a Convolutional Neural Network to calculate IC[Formula: see text] values for every combination of cell lines and drugs. Then a threshold IC[Formula: see text] value is obtained for each drug by performing K-means clustering of their known IC[Formula: see text] values. Finally, with the help of this threshold value, cell lines are classified as either sensitive or resistant to each drug. Results: Experimental results indicate that AGCN has an accuracy of 0.82 and performs better than many existing methods. In addition to that, we have done external validation of AGCN using data taken from The Cancer Genome Atlas (TCGA) clinical database, and we got an accuracy of 0.91. Conclusion: According to the results obtained, concatenating multi-omics data with drug structure data using AGCN for drug response prediction tasks greatly improves the accuracy of the prediction task.

Predicting locus-specific DNA methylation levels in cancer and paracancer tissues

Aim: To predict base-resolution DNA methylation in cancerous and paracancerous tissues. Material & methods: We collected six cancer DNA methylation datasets from The Cancer Genome Atlas and five cancer datasets from Gene Expression Omnibus and established machine learning models using paired cancerous and paracancerous tissues. Tenfold cross-validation and independent validation were performed to demonstrate the effectiveness of the proposed method. Results: The developed cross-tissue prediction models can substantially increase the accuracy at more than 68% of CpG sites and contribute to enhancing the statistical power of differential methylation analyses. An XGBoost model leveraging multiple correlating CpGs may elevate the prediction accuracy. Conclusion: This study provides a powerful tool for DNA methylation analysis and has the potential to gain new insights into cancer research from epigenetics.

Interaction molecular QTL mapping discovers cellular and environmental modifiers of genetic regulatory effects

Bulk-tissue molecular quantitative trait loci (QTLs) have been the starting point for interpreting disease-associated variants, and context-specific QTLs show particular relevance for disease. Here, we present the results of mapping interaction QTLs (iQTLs) for cell type, age, and other phenotypic variables in multi-omic, longitudinal data from the blood of individuals of diverse ancestries. By modeling the interaction between genotype and estimated cell-type proportions, we demonstrate that cell-type iQTLs could be considered as proxies for cell-type-specific QTL effects, particularly for the most abundant cell type in the tissue. The interpretation of age iQTLs, however, warrants caution because the moderation effect of age on the genotype and molecular phenotype association could be mediated by changes in cell-type composition. Finally, we show that cell-type iQTLs contribute to cell-type-specific enrichment of diseases that, in combination with additional functional data, could guide future functional studies. Overall, this study highlights the use of iQTLs to gain insights into the context specificity of regulatory effects.

MLACNN: an attention mechanism-based CNN architecture for predicting genome-wide DNA methylation

Methylation is an important epigenetic regulation of methylation genes that plays a crucial role in regulating biological processes. While traditional methods for detecting methylation in biological experiments are constantly improving, the development of artificial intelligence has led to the emergence of deep learning and machine learning methods as a new trend. However, traditional machine learning-based methods rely heavily on manual feature extraction, and most deep learning methods for studying methylation extract fewer features due to their simple network structures. To address this, we propose a bottomneck network based on an attention mechanism and use new methods to ensure that the deep network can learn more effective features while minimizing overfitting. This approach enables the model to learn more features from nucleotide sequences and make better predictions of methylation. The model uses three coding methods to encode the original DNA sequence and then applies feature fusion based on attention mechanisms to obtain the best fusion method. Our results demonstrate that MLACNN outperforms previous methods and achieves more satisfactory performance.

DeepMethylation: a deep learning based framework with GloVe and Transformer encoder for DNA methylation prediction

DNA methylation is a crucial topic in bioinformatics research. Traditional wet experiments are usually time-consuming and expensive. In contrast, machine learning offers an efficient and novel approach. In this study, we propose DeepMethylation, a novel methylation predictor with deep learning. Specifically, the DNA sequence is encoded with word embedding and GloVe in the first step. After that, dilated convolution and Transformer encoder are utilized to extract the features. Finally, full connection and softmax operators are applied to predict the methylation sites. The proposed model achieves an accuracy of 97.8% on the 5mC dataset, which outperforms state-of-the-art methods. Furthermore, our predictor exhibits good generalization ability as it achieves an accuracy of 95.8% on the m1A dataset. To ease access for other researchers, our code is publicly available at https://github.com/sb111169/tf-5mc.

Alterations in Growth Habit to Channel End-of-Season Perennial Reserves towards Increased Yield and Reduced Regrowth after Defoliation in Upland Cotton (Gossypium hirsutum L.)

Cotton (Gossypium spp.) is the primary source of natural textile fiber in the U.S. and a major crop in the Southeastern U.S. Despite constant efforts to increase the cotton fiber yield, the yield gain has stagnated. Therefore, we undertook a novel approach to improve the cotton fiber yield by altering its growth habit from perennial to annual. In this effort, we identified genotypes with high-expression alleles of five floral induction and meristem identity genes (FT, SOC1, FUL, LFY, and AP1) from an Upland cotton mini-core collection and crossed them in various combinations to develop cotton lines with annual growth habit, optimal flowering time, and enhanced productivity. To facilitate the characterization of genotypes with the desired combinations of stacked alleles, we identified molecular markers associated with the gene expression traits via genome-wide association analysis using a 63 K SNP Array. Over 14,500 SNPs showed polymorphism and were used for association analysis. A total of 396 markers showed associations with expression traits. Of these 396 markers, 159 were mapped to genes, 50 to untranslated regions, and 187 to random genomic regions. Biased genomic distribution of associated markers was observed where more trait-associated markers mapped to the cotton D sub-genome. Many quantitative trait loci coincided at specific genomic regions. This observation has implications as these traits could be bred together. The analysis also allowed the identification of candidate regulators of the expression patterns of these floral induction and meristem identity genes whose functions will be validated.

Towards understanding paleoclimate impacts on primate de novo genes

De novo genes are genes that emerge as new genes in some species, such as primate de novo genes that emerge in certain primate species. Over the past decade, a great deal of research has been conducted regarding their emergence, origins, functions, and various attributes in different species, some of which have involved estimating the ages of de novo genes. However, limited by the number of species available for whole-genome sequencing, relatively few studies have focused specifically on the emergence time of primate de novo genes. Among those, even fewer investigate the association between primate gene emergence with environmental factors, such as paleoclimate (ancient climate) conditions. This study investigates the relationship between paleoclimate and human gene emergence at primate species divergence. Based on 32 available primate genome sequences, this study has revealed possible associations between temperature changes and the emergence of de novo primate genes. Overall, findings in this study are that de novo genes tended to emerge in the recent 13 MY when the temperature continues cooling, which is consistent with past findings. Furthermore, in the context of an overall trend of cooling temperature, new primate genes were more likely to emerge during local warming periods, where the warm temperature more closely resembled the environmental condition that preceded the cooling trend. Results also indicate that both primate de novo genes and human cancer-associated genes have later origins in comparison to random human genes. Future studies can be in-depth on understanding human de novo gene emergence from an environmental perspective as well as understanding species divergence from a gene emergence perspective.

iCpG-Pos: an accurate computational approach for identification of CpG sites using positional features on single-cell whole genome sequence data

MOTIVATION

The investigation of DNA methylation can shed light on the processes underlying human well-being and help determine overall human health. However, insufficient coverage makes it challenging to implement single-stranded DNA methylation sequencing technologies, highlighting the need for an efficient prediction model. Models are required to create an understanding of the underlying biological systems and to project single-cell (methylated) data accurately.

RESULTS

In this study, we developed positional features for predicting CpG sites. Positional characteristics of the sequence are derived using data from CpG regions and the separation between nearby CpG sites. Multiple optimized classifiers and different ensemble learning approaches are evaluated. The OPTUNA framework is used to optimize the algorithms. The CatBoost algorithm followed by the stacking algorithm outperformed existing DNA methylation identifiers.

AVAILABILITY AND IMPLEMENTATION

The data and methodologies used in this study are openly accessible to the research community. Researchers can access the positional features and algorithms used for predicting CpG site methylation patterns. To achieve superior performance, we employed the CatBoost algorithm followed by the stacking algorithm, which outperformed existing DNA methylation identifiers. The proposed iCpG-Pos approach utilizes only positional features, resulting in a substantial reduction in computational complexity compared to other known approaches for detecting CpG site methylation patterns. In conclusion, our study introduces a novel approach, iCpG-Pos, for predicting CpG site methylation patterns. By focusing on positional features, our model offers both accuracy and efficiency, making it a promising tool for advancing DNA methylation research and its applications in human health and well-being.

Interaction molecular QTL mapping discovers cellular and environmental modifiers of genetic regulatory effects

Bulk tissue molecular quantitative trait loci (QTLs) have been the starting point for interpreting disease-associated variants, while context-specific QTLs show particular relevance for disease. Here, we present the results of mapping interaction QTLs (iQTLs) for cell type, age, and other phenotypic variables in multi-omic, longitudinal data from blood of individuals of diverse ancestries. By modeling the interaction between genotype and estimated cell type proportions, we demonstrate that cell type iQTLs could be considered as proxies for cell type-specific QTL effects. The interpretation of age iQTLs, however, warrants caution as the moderation effect of age on the genotype and molecular phenotype association may be mediated by changes in cell type composition. Finally, we show that cell type iQTLs contribute to cell type-specific enrichment of diseases that, in combination with additional functional data, may guide future functional studies. Overall, this study highlights iQTLs to gain insights into the context-specificity of regulatory effects.

On the prediction of non-CG DNA methylation using machine learning

DNA methylation can be detected and measured using sequencing instruments after sodium bisulfite conversion, but experiments can be expensive for large eukaryotic genomes. Sequencing nonuniformity and mapping biases can leave parts of the genome with low or no coverage, thus hampering the ability of obtaining DNA methylation levels for all cytosines. To address these limitations, several computational methods have been proposed that can predict DNA methylation from the DNA sequence around the cytosine or from the methylation level of nearby cytosines. However, most of these methods are entirely focused on CG methylation in humans and other mammals. In this work, we study, for the first time, the problem of predicting cytosine methylation for CG, CHG and CHH contexts on six plant species, either from the DNA primary sequence around the cytosine or from the methylation levels of neighboring cytosines. In this framework, we also study the cross-species prediction problem and the cross-context prediction problem (within the same species). Finally, we show that providing gene and repeat annotations allows existing classifiers to significantly improve their prediction accuracy. We introduce a new classifier called AMPS (annotation-based methylation prediction from sequence) that takes advantage of genomic annotations to achieve higher accuracy.

Recall DNA methylation levels at low coverage sites using a CNN model in WGBS

DNA methylation is an important regulator of gene transcription. WGBS is the gold-standard approach for base-pair resolution quantitative of DNA methylation. It requires high sequencing depth. Many CpG sites with insufficient coverage in the WGBS data, resulting in inaccurate DNA methylation levels of individual sites. Many state-of-arts computation methods were proposed to predict the missing value. However, many methods required either other omics datasets or other cross-sample data. And most of them only predicted the state of DNA methylation. In this study, we proposed the RcWGBS, which can impute the missing (or low coverage) values from the DNA methylation levels on the adjacent sides. Deep learning techniques were employed for the accurate prediction. The WGBS datasets of H1-hESC and GM12878 were down-sampled. The average difference between the DNA methylation level at 12× depth predicted by RcWGBS and that at >50× depth in the H1-hESC and GM2878 cells are less than 0.03 and 0.01, respectively. RcWGBS performed better than METHimpute even though the sequencing depth was as low as 12×. Our work would help to process methylation data of low sequencing depth. It is beneficial for researchers to save sequencing costs and improve data utilization through computational methods.

Inferring CpG methylation signatures accumulated along human history from genetic variation catalogs

Understanding the DNA methylation patterns in the human genome is a key step to decipher gene regulatory mechanisms and model mutation rate heterogeneity in the human genome. While methylation rates can be measured e.g. with bisulfite sequencing, such measures do not capture historical patterns. Here we present a new method, Methylation Hidden Markov Model (MHMM), to estimate the accumulated germline methylation signature in human population history leveraging two properties: (1) Mutation rates of cytosine to thymine transitions at methylated CG dinucleotides are orders of magnitude higher than that in the rest of the genome. (2) Methylation levels are locally correlated, so the allele frequencies of neighboring CpGs can be used jointly to estimate methylation status. We applied MHMM to allele frequencies from the TOPMed and the gnomAD genetic variation catalogs. Our estimates are consistent with whole genome bisulfite sequencing (WGBS) measured human germ cell methylation levels at 90% of CpG sites, but we also identified ~ 442, 000 historically methylated CpG sites that could not be captured due to sample genetic variation, and inferred methylation status for ~ 721, 000 CpG sites that were missing from WGBS. Hypo-methylated regions identified by combining our results with experimental measures are 1.7 times more likely to recover known active genomic regions than those identified by WGBS alone. Our estimated historical methylation status can be leveraged to enhance bioinformatic analysis of germline methylation such as annotating regulatory and inactivated genomic regions and provide insights in sequence evolution including predicting mutation constraint.

DNA methylation-based profiling of horse archaeological remains for age-at-death and castration

Age profiling of archaeological bone assemblages can inform on past animal management practices, but is limited by the fragmentary nature of the fossil record and the lack of universal skeletal markers for age. DNA methylation clocks offer new, albeit challenging, alternatives for estimating the age-at-death of ancient individuals. Here, we take advantage of the availability of a DNA methylation clock based on 31,836 CpG sites and dental age markers in horses to assess age predictions in 84 ancient remains. We evaluate our approach using whole-genome sequencing data and develop a capture assay providing reliable estimates for only a fraction of the cost. We also leverage DNA methylation patterns to assess castration practice in the past. Our work opens for a deeper characterization of past husbandry and ritual practices and holds the potential to reveal age mortality profiles in ancient societies, once extended to human remains.

The impact of spatial correlation on methylation entropy with application to mouse brain methylome

BACKGROUND

With the advance of bisulfite sequencing technologies, massive amount of methylation data have been generated, which provide unprecedented opportunities to study the epigenetic mechanism and its relationship to other biological processes. A commonly seen feature of the methylation data is the correlation between nearby CpG sites. Although such a spatial correlation was utilized in several epigenetic studies, its interaction to other characteristics of the methylation data has not been fully investigated.

RESULTS

We filled this research gap from an information theoretic perspective, by exploring the impact of the spatial correlation on the methylation entropy (ME). With the spatial correlation taken into account, we derived the analytical relation between the ME and another key parameter, the methylation probability. By comparing it to the empirical relation between the two corresponding statistics, the observed ME and the mean methylation level, genomic loci under strong epigenetic control can be identified, which may serve as potential markers for cell-type specific methylation. The proposed method was validated by simulation studies, and applied to analyze a published dataset of mouse brain methylome.

CONCLUSIONS

Compared to other sophisticated methods developed in literature, the proposed method provides a simple but effective way to detect CpG segments under strong epigenetic control (e.g., with bipolar methylation pattern). Findings from this study shed light on the identification of cell-type specific genes/pathways based on methylation data from a mixed cell population.

Computational Methods for Single-cell DNA Methylome Analysis

Dissecting intercellular epigenetic differences is key to understanding tissue heterogeneity. Recent advances in single-cell DNA methylome profiling have presented opportunities to resolve this heterogeneity at the maximum resolution. While these advances enable us to explore frontiers of chromatin biology and better understand cell lineage relationships, they pose new challenges in data processing and interpretation. This review surveys the current state of computational tools developed for single-cell DNA methylome data analysis. We discuss critical components of single-cell DNA methylome data analysis, including data preprocessing, quality control, imputation, dimensionality reduction, cell clustering, supervised cell annotation, cell lineage reconstruction, gene activity scoring, and integration with transcriptome data. We also highlight unique aspects of single-cell DNA methylome data analysis and discuss how techniques common to other single-cell omics data analyses can be adapted to analyze DNA methylomes. Finally, we discuss existing challenges and opportunities for future development.

Comparative analysis of genome-scale, base-resolution DNA methylation profiles across 580 animal species

Methylation of cytosines is a prototypic epigenetic modification of the DNA. It has been implicated in various regulatory mechanisms across the animal kingdom and particularly in vertebrates. We mapped DNA methylation in 580 animal species (535 vertebrates, 45 invertebrates), resulting in 2443 genome-scale DNA methylation profiles of multiple organs. Bioinformatic analysis of this large dataset quantified the association of DNA methylation with the underlying genomic DNA sequence throughout vertebrate evolution. We observed a broadly conserved link with two major transitions-once in the first vertebrates and again with the emergence of reptiles. Cross-species comparisons focusing on individual organs supported a deeply conserved association of DNA methylation with tissue type, and cross-mapping analysis of DNA methylation at gene promoters revealed evolutionary changes for orthologous genes. In summary, this study establishes a large resource of vertebrate and invertebrate DNA methylomes, it showcases the power of reference-free epigenome analysis in species for which no reference genomes are available, and it contributes an epigenetic perspective to the study of vertebrate evolution.

DNA methylation QTL mapping across diverse human tissues provides molecular links between genetic variation and complex traits

Studies of DNA methylation (DNAm) in solid human tissues are relatively scarce; tissue-specific characterization of DNAm is needed to understand its role in gene regulation and its relevance to complex traits. We generated array-based DNAm profiles for 987 human samples from the Genotype-Tissue Expression (GTEx) project, representing 9 tissue types and 424 subjects. We characterized methylome and transcriptome correlations (eQTMs), genetic regulation in cis (mQTLs and eQTLs) across tissues and e/mQTLs links to complex traits. We identified mQTLs for 286,152 CpG sites, many of which (>5%) show tissue specificity, and mQTL colocalizations with 2,254 distinct GWAS hits across 83 traits. For 91% of these loci, a candidate gene link was identified by integration of functional maps, including eQTMs, and/or eQTL colocalization, but only 33% of loci involved an eQTL and mQTL present in the same tissue type. With this DNAm-focused integrative analysis, we contribute to the understanding of molecular regulatory mechanisms in human tissues and their impact on complex traits.

A comparison of the genes and genesets identified by GWAS and EWAS of fifteen complex traits

Identifying genomic regions pertinent to complex traits is a common goal of genome-wide and epigenome-wide association studies (GWAS and EWAS). GWAS identify causal genetic variants, directly or via linkage disequilibrium, and EWAS identify variation in DNA methylation associated with a trait. While GWAS in principle will only detect variants due to causal genes, EWAS can also identify genes via confounding, or reverse causation. We systematically compare GWAS (N > 50,000) and EWAS (N > 4500) results of 15 complex traits. We evaluate if the genes or gene ontology terms flagged by GWAS and EWAS overlap, and find substantial overlap for diastolic blood pressure, (gene overlap P = 5.2 × 10^-6; term overlap P = 0.001). We superimpose our empirical findings against simulated models of varying genetic and epigenetic architectures and observe that in most cases GWAS and EWAS are likely capturing distinct genesets. Our results indicate that GWAS and EWAS are capturing different aspects of the biology of complex traits.

Investigating the prediction of CpG methylation levels from SNP genotype data to help elucidate relationships between methylation, gene expression and complex traits

As popularised by PrediXcan (and related methods), transcriptome-wide association studies (TWAS), in which gene expression is imputed from single-nucleotide polymorphism (SNP) genotypes and tested for association with a phenotype, are a popular approach for investigating the role of gene expression in complex traits. Like gene expression, DNA methylation is an important biological process and, being under genetic regulation, may be imputable from SNP genotypes. Here, we investigate prediction of CpG methylation levels from SNP genotype data to help elucidate relationships between methylation, gene expression and complex traits. We start by examining how well CpG methylation can be predicted from SNP genotypes, comparing three penalised regression approaches and examining whether changing the window size improves prediction accuracy. Although methylation at most CpG sites cannot be accurately predicted from SNP genotypes, for a subset it can be predicted well. We next apply our methylation prediction models (trained using the optimal method and window size) to carry out a methylome-wide association study (MWAS) of primary biliary cholangitis. We intersect the regions identified via MWAS with those identified via TWAS, providing insight into the interplay between CpG methylation, gene expression and disease status. We conclude that MWAS has the potential to improve understanding of biological mechanisms in complex traits.

Comparative analysis of Illumina Mouse Methylation BeadChip and reduced-representation bisulfite sequencing for routine DNA methylation analysis

Researching the murine epigenome in disease models has been hampered by the lack of appropriate and cost-effective DNA methylation arrays. Here we perform a comprehensive, comparative analysis between the Mouse Methylation BeadChip (MMB) and reduced-representation bisulfite sequencing (RRBS) in two murine models of colorectal carcinogenesis. We evaluate the coverage, variability, and ability to identify differential DNA methylation of RRBS and MMB. We show that MMB is an effective tool for profiling the murine methylome that performs comparably with RRBS, identifying similar differentially methylated pathways. Although choice of technology is experiment dependent and will be predicated on the underlying biology being probed, these analyses provide insights into the relative strengths and weaknesses of each approach.

Completing Single-Cell DNA Methylome Profiles via Transfer Learning Together With KL-Divergence

The high level of sparsity in methylome profiles obtained using whole-genome bisulfite sequencing in the case of low biological material amount limits its value in the study of systems in which large samples are difficult to assemble, such as mammalian preimplantation embryonic development. The recently developed computational methods for addressing the sparsity by imputing missing have their limits when the required minimum data coverage or profiles of the same tissue in other modalities are not available. In this study, we explored the use of transfer learning together with Kullback-Leibler (KL) divergence to train predictive models for completing methylome profiles with very low coverage (below 2%). Transfer learning was used to leverage less sparse profiles that are typically available for different tissues for the same species, while KL divergence was employed to maximize the usage of information carried in the input data. A deep neural network was adopted to extract both DNA sequence and local methylation patterns for imputation. Our study of training models for completing methylome profiles of bovine oocytes and early embryos demonstrates the effectiveness of transfer learning and KL divergence, with individual increase of 29.98 and 29.43%, respectively, in prediction performance and 38.70% increase when the two were used together. The drastically increased data coverage (43.80-73.6%) after imputation powers downstream analyses involving methylomes that cannot be effectively done using the very low coverage profiles (0.06-1.47%) before imputation.

Adipose methylome integrative-omic analyses reveal genetic and dietary metabolic health drivers and insulin resistance classifiers

BACKGROUND

There is considerable evidence for the importance of the DNA methylome in metabolic health, for example, a robust methylation signature has been associated with body mass index (BMI). However, visceral fat (VF) mass accumulation is a greater risk factor for metabolic disease than BMI alone. In this study, we dissect the subcutaneous adipose tissue (SAT) methylome signature relevant to metabolic health by focusing on VF as the major risk factor of metabolic disease. We integrate results with genetic, blood methylation, SAT gene expression, blood metabolomic, dietary intake and metabolic phenotype data to assess and quantify genetic and environmental drivers of the identified signals, as well as their potential functional roles.

METHODS

Epigenome-wide association analyses were carried out to determine visceral fat mass-associated differentially methylated positions (VF-DMPs) in SAT samples from 538 TwinsUK participants. Validation and replication were performed in 333 individuals from 3 independent cohorts. To assess functional impacts of the VF-DMPs, the association between VF and gene expression was determined at the genes annotated to the VF-DMPs and an association analysis was carried out to determine whether methylation at the VF-DMPs is associated with gene expression. Further epigenetic analyses were carried out to compare methylation levels at the VF-DMPs as the response variables and a range of different metabolic health phenotypes including android:gynoid fat ratio (AGR), lipids, blood metabolomic profiles, insulin resistance, T2D and dietary intake variables. The results from all analyses were integrated to identify signals that exhibit altered SAT function and have strong relevance to metabolic health.

RESULTS

We identified 1181 CpG positions in 788 genes to be differentially methylated with VF (VF-DMPs) with significant enrichment in the insulin signalling pathway. Follow-up cross-omic analysis of VF-DMPs integrating genetics, gene expression, metabolomics, diet, and metabolic traits highlighted VF-DMPs located in 9 genes with strong relevance to metabolic disease mechanisms, with replication of signals in FASN, SREBF1, TAGLN2, PC and CFAP410. PC methylation showed evidence for mediating effects of diet on VF. FASN DNA methylation exhibited putative causal effects on VF that were also strongly associated with insulin resistance and methylation levels in FASN better classified insulin resistance (AUC=0.91) than BMI or VF alone.

CONCLUSIONS

Our findings help characterise the adiposity-associated methylation signature of SAT, with insights for metabolic disease risk.

mRNA expression and epigenetic-based role of chromodomain helicase DNA-binding 5 in hepatocellular carcinoma

Objective

Chromodomain helicase DNA-binding 5 (CHD5) acts as a tumor suppressor gene in some cancers. CHD5 expression levels may affect an individual’s susceptibility to hepatocellular carcinoma (HCC). This study aimed to evaluate the methylation pattern of the CHD5 promoter region and the gene’s corresponding mRNA expression in HCC patients compared with healthy individuals.

Methods

In this case–control study, CHD5 mRNA gene expression levels and DNA methylation patterns were analyzed in 81 HCC patients and 90 healthy individuals by quantitative reverse transcription polymerase chain reaction and methylation-specific polymerase chain reaction, respectively.

Results

The CHD5 gene was hypermethylated in 61.8% of the HCC patients and 54.4% of the controls, and this difference was statistically significant. The CHD5 mRNA expression levels were significantly lower in the HCC patient group.

Conclusions

Hypermethylation of the CHD5 promoter region may significantly lower the expression of this gene, affecting the incidence and severity of HCC. The methylation status of CHD5 can also be further studied as a prognostic factor in HCC.

Association between DNA methylation variability and self-reported exposure to heavy metals

Individuals encounter varying environmental exposures throughout their lifetimes. Some exposures such as smoking are readily observed and have high personal recall; others are more indirect or sporadic and might only be inferred from long occupational histories or lifestyles. We evaluated the utility of using lifetime-long self-reported exposures for identifying differential methylation in an amyotrophic lateral sclerosis cases-control cohort of 855 individuals. Individuals submitted paper-based surveys on exposure and occupational histories as well as whole blood samples. Genome-wide DNA methylation levels were quantified using the Illumina Infinium Human Methylation450 array. We analyzed 15 environmental exposures using the OSCA software linear and MOA models, where we regressed exposures individually by methylation adjusted for batch effects and disease status as well as predicted scores for age, sex, cell count, and smoking status. We also regressed on the first principal components on clustered environmental exposures to detect DNA methylation changes associated with a more generalised definition of environmental exposure. Five DNA methylation probes across three environmental exposures (cadmium, mercury and metalwork) were significantly associated using the MOA models and seven through the linear models, with one additionally across a principal component representing chemical exposures. Methylome-wide significance for four of these markers was driven by extreme hyper/hypo-methylation in small numbers of individuals. The results indicate the potential for using self-reported exposure histories in detecting DNA methylation changes in response to the environment, but also highlight the confounded nature of environmental exposure in cohort studies.

The Epigenetic Mechanisms Involved in Chronic Pain in Rodents: A Mini- Review

Chronic pain is a common distressing neurological disorder and about 30% of the global population suffers from it. In addition to being highly prevalent, chronic pain causes a heavy economic and social burden. Although substantial progress has been achieved to dissect the underlying mechanism of chronic pain in the past few decades, the incidence and treatment of this neurological illness is yet not properly managed in clinical practice. While nerve injury-, chemotherapy- or inflammation-induced functional regulation of gene expression in the dorsal root ganglion and spinal cord are extensively reported to be involved in the pathogenic process of chronic pain, the specific mechanism of these altered transcriptional profile still remains unclear. Recent studies have shown that epigenetic mechanisms, including DNA/RNA methylation, histone modification and circular RNAs regulation, are involved in the occurrence and development of chronic pain. In this review, we provide a description of research on the role of epigenetic mechanism in chronic pain, summarize the latest clinical and preclinical advance in this field, and propose the potential directions for further research to elucidate the molecular mechanism underlying the pathogenesis of chronic pain.

Lycophyte transcriptomes reveal two whole-genome duplications in Lycopodiaceae: Insights into the polyploidization of Phlegmariurus

Lycophytes are an ancient clade of the non-flowering vascular plants with chromosome numbers that vary from tens to hundreds. They are an excellent study system for examining whole-genome duplications (WGDs), or polyploidization, in spore-dispersed vascular plants. However, a lack of genome sequence data limits the reliable detection of very ancient WGDs, small-scale duplications (SSDs), and recent WGDs. Here, we integrated phylogenomic analysis and the distribution of synonymous substitutions per synonymous sites (Ks) of the transcriptomes of 13 species of lycophytes to identify, locate, and date multiple WGDs in the lycophyte family Lycopodiaceae. Additionally, we examined the genus Phlegmariurus for signs of genetic discordance, which can provide valuable insight into the underlying causes of such conflict (e.g., hybridization, incomplete lineage sorting, or horizontal gene transfer).We found strong evidence that two WGD events occurred along the phylogenetic backbone of Lycopodiaceae, with one occurring in the common ancestor of extant Phlegmariurus (Lycopodiaceae) approximately 22-23 million years ago (Mya) and the other occurring in the common ancestor of Lycopodiaceae around 206-214 Mya. Interestingly, we found significant genetic discordance in the genus Phlegmariurus, indicating that the genus has a complex evolutionary history. This study provides molecular evidence for multiple WGDs in Lycopodiaceae and offers phylogenetic clues to the evolutionary history of Lycopodiaceae.

Predicting male fertility from the sperm methylome: application to 120 bulls with hundreds of artificial insemination records

BACKGROUND

Conflicting results regarding alterations to sperm DNA methylation in cases of spermatogenesis defects, male infertility and poor developmental outcomes have been reported in humans. Bulls used for artificial insemination represent a relevant model in this field, as the broad dissemination of bull semen considerably alleviates confounding factors and enables the precise assessment of male fertility. This study was therefore designed to assess the potential for sperm DNA methylation to predict bull fertility.

RESULTS

A unique collection of 100 sperm samples was constituted by pooling 2-5 ejaculates per bull from 100 Montbéliarde bulls of comparable ages, assessed as fertile (n = 57) or subfertile (n = 43) based on non-return rates 56 days after insemination. The DNA methylation profiles of these samples were obtained using reduced representation bisulfite sequencing. After excluding putative sequence polymorphisms, 490 fertility-related differentially methylated cytosines (DMCs) were identified, most of which were hypermethylated in subfertile bulls. Interestingly, 46 genes targeted by DMCs are involved in embryonic and fetal development, sperm function and maturation, or have been related to fertility in genome-wide association studies; five of these were further analyzed by pyrosequencing. In order to evaluate the prognostic value of fertility-related DMCs, the sperm samples were split between training (n = 67) and testing (n = 33) sets. Using a Random Forest approach, a predictive model was built from the methylation values obtained on the training set. The predictive accuracy of this model was 72% on the testing set and 72% on individual ejaculates collected from an independent cohort of 20 bulls.

CONCLUSION

This study, conducted on the largest set of bull sperm samples so far examined in epigenetic analyses, demonstrated that the sperm methylome is a valuable source of male fertility biomarkers. The next challenge is to combine these results with other data on the same sperm samples in order to improve the quality of the model and better understand the interplay between DNA methylation and other molecular features in the regulation of fertility. This research may have potential applications in human medicine, where infertility affects the interaction between a male and a female, thus making it difficult to isolate the male factor.

Targeting of Ubiquitin E3 Ligase RNF5 as a Novel Therapeutic Strategy in Neuroectodermal Tumors

RNF5, an endoplasmic reticulum (ER) E3 ubiquitin ligase, participates to the ER-associated protein degradation guaranteeing the protein homeostasis. Depending on tumor model tested, RNF5 exerts pro- or anti-tumor activity. The aim of this study was to elucidate the controversial role of RNF5 in neuroblastoma and melanoma, two neuroectodermal tumors of infancy and adulthood, respectively. RNF5 gene levels are evaluated in publicly available datasets reporting the gene expression profile of melanoma and neuroblastoma primary tumors at diagnosis. The therapeutic effect of Analog-1, an RNF5 pharmacological activator, was investigated on in vitro and in vivo neuroblastoma and melanoma models. In both neuroblastoma and melanoma patients the high expression of RNF5 correlated with a better prognostic outcome. Treatment of neuroblastoma and melanoma cell lines with Analog-1 reduced cell viability by impairing the glutamine availability and energy metabolism through inhibition of F1Fo ATP-synthase activity. This latter event led to a marked increase in oxidative stress, which, in turn, caused cell death. Similarly, neuroblastoma- and melanoma-bearing mice treated with Analog-1 showed a significant delay of tumor growth in comparison to those treated with vehicle only. These findings validate RNF5 as an innovative drug target and support the development of Analog-1 in early phase clinical trials for neuroblastoma and melanoma patients.

GMQN: A Reference-Based Method for Correcting Batch Effects and Probe Bias in HumanMethylation BeadChip

The Illumina HumanMethylation BeadChip is one of the most cost-effective methods to quantify DNA methylation levels at single-base resolution across the human genome, which makes it a routine platform for epigenome-wide association studies. It has accumulated tens of thousands of DNA methylation array samples in public databases, providing great support for data integration and further analysis. However, the majority of public DNA methylation data are deposited as processed data without background probes which are widely used in data normalization. Here, we present Gaussian mixture quantile normalization (GMQN), a reference based method for correcting batch effects as well as probe bias in the HumanMethylation BeadChip. Availability and implementation: https://github.com/MengweiLi-project/gmqn.

Accurate Measurement of DNA Methylation: Challenges and Bias Correction

DNA methylation is a key epigenetic modification involved in gene regulation whose contribution to disease susceptibility is still not fully understood. As the cost of genome sequencing technologies continues to drop, it will soon become commonplace to perform genome-wide quantification of DNA methylation at a single base-pair resolution. However, the demand for its accurate quantification might vary across studies. When the scope of the analysis is to detect regions of the genome with different methylation patterns between two or more conditions, e.g., case vs control; treatments vs placebo, accuracy is not crucial. This is the case in epigenome-wide association studies used as genome-wide screening of methylation changes to detect new candidate genes and regions associated with a specific disease or condition. If the aim of the analysis is to use DNA methylation measurements as a biomarker for diseases diagnosis and treatment (Laird, Nat Rev Cancer 3:253-266, 2003; Bock, Epigenomics 1:99-110, 2009), it is instead recommended to produce accurate methylation measurements. Furthermore, if the objective is the detection of DNA methylation in subclonal tumor cell populations or in circulating tumor DNA or in any case of mosaicism, the importance of accuracy becomes critical. The aim of this chapter is to describe the factors that could affect the precise measurement of methylation levels and a recent Bayesian statistical method called MethylCal and its extension that have been proposed to minimize this problem.

DNA Methylation Imputation Across Platforms

In this chapter, we will provide a review on imputation in the context of DNA methylation, specifically focusing on a penalized functional regression (PFR) method we have previously developed. We will start with a brief review of DNA methylation, genomic and epigenomic contexts where imputation has proven beneficial in practice, and statistical or computational methods proposed for DNA methylation in the recent literature (Subheading 1). The rest of the chapter (Subheadings 2-4) will provide a detailed review of our PFR method proposed for across-platform imputation, which incorporates nonlocal information using a penalized functional regression framework. Subheading 2 introduces commonly employed technologies for DNA methylation measurement and describes the real dataset we have used in the development of our method: the acute myeloid leukemia (AML) dataset from The Cancer Genome Atlas (TCGA) project. Subheading 3 comprehensively reviews our method, encompassing data harmonization prior to model building, the actual building of penalized functional regression model, post-imputation quality filter, and imputation quality assessment. Subheading 4 shows the performance of our method in both simulation and the TCGA AML dataset, demonstrating that our penalized functional regression model is a valuable across-platform imputation tool for DNA methylation data, particularly because of its ability to boost statistical power for subsequent epigenome-wide association study. Finally, Subheading 5 provides future perspectives on imputation for DNA methylation data.

CpG Transformer for imputation of single-cell methylomes

MOTIVATION

The adoption of current single-cell DNA methylation sequencing protocols is hindered by incomplete coverage, outlining the need for effective imputation techniques. The task of imputing single-cell (methylation) data requires models to build an understanding of underlying biological processes.

RESULTS

We adapt the transformer neural network architecture to operate on methylation matrices through combining axial attention with sliding window self-attention. The obtained CpG Transformer displays state-of-the-art performances on a wide range of scBS-seq and scRRBS-seq datasets. Furthermore, we demonstrate the interpretability of CpG Transformer and illustrate its rapid transfer learning properties, allowing practitioners to train models on new datasets with a limited computational and time budget.

AVAILABILITY AND IMPLEMENTATION

CpG Transformer is freely available at https://github.com/gdewael/cpg-transformer.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Machine learning applications for therapeutic tasks with genomics data

Thanks to the increasing availability of genomics and other biomedical data, many machine learning algorithms have been proposed for a wide range of therapeutic discovery and development tasks. In this survey, we review the literature on machine learning applications for genomics through the lens of therapeutic development. We investigate the interplay among genomics, compounds, proteins, electronic health records, cellular images, and clinical texts. We identify 22 machine learning in genomics applications that span the whole therapeutics pipeline, from discovering novel targets, personalizing medicine, developing gene-editing tools, all the way to facilitating clinical trials and post-market studies. We also pinpoint seven key challenges in this field with potentials for expansion and impact. This survey examines recent research at the intersection of machine learning, genomics, and therapeutic development.

Pharmacologic approaches to amino acid depletion for cancer therapy

Cancer cells undergo metabolic reprogramming to support increased demands in bioenergetics and biosynthesis and to maintain reactive oxygen species at optimum levels. As metabolic alterations are broadly observed across many cancer types, metabolic reprogramming is considered a hallmark of cancer. A metabolic alteration commonly seen in cancer cells is an increased demand for certain amino acids. Amino acids are involved in a wide range of cellular functions, including proliferation, redox balance, bioenergetic and biosynthesis support, and homeostatic functions. Thus, targeting amino acid dependency in cancer is an attractive strategy for a number of cancers. In particular, pharmacologically mediated amino acid depletion has been evaluated as a cancer treatment option for several cancers. Amino acids that have been investigated for the feasibility of drug-induced depletion in preclinical and clinical studies for cancer treatment include arginine, asparagine, cysteine, glutamine, lysine, and methionine. In this review, we will summarize the status of current research on pharmacologically mediated amino acid depletion as a strategy for cancer treatment and potential chemotherapeutic combinations that synergize with amino acid depletion to further inhibit tumor growth and progression.

Synthetic biology of plant natural products: From pathway elucidation to engineered biosynthesis in plant cells

Plant natural products (PNPs) are the main sources of drugs, food additives, and new biofuels and have become a hotspot in synthetic biology. In the past two decades, the engineered biosynthesis of many PNPs has been achieved through the construction of microbial cell factories. Alongside the rapid development of plant physiology, genetics, and plant genetic modification techniques, hosts have now expanded from single-celled microbes to complex plant systems. Plant synthetic biology is an emerging field that combines engineering principles with plant biology. In this review, we introduce recent advances in the biosynthetic pathway elucidation of PNPs and summarize the progress of engineered PNP biosynthesis in plant cells. Furthermore, a future vision of plant synthetic biology is proposed. Although we are still a long way from overcoming all the bottlenecks in plant synthetic biology, the ascent of this field is expected to provide a huge opportunity for future agriculture and industry.

Prenatal risk factors and neonatal DNA methylation in very preterm infants

Background

Prenatal risk factors are related to poor health and developmental outcomes for infants, potentially via epigenetic mechanisms. We tested associations between person-centered prenatal risk profiles, cumulative prenatal risk models, and epigenome-wide DNA methylation (DNAm) in very preterm neonates.

Methods

We studied 542 infants from a multi-center study of infants born < 30 weeks postmenstrual age. We assessed 24 prenatal risk factors via maternal report and medical record review. Latent class analysis was used to define prenatal risk profiles. DNAm was quantified from neonatal buccal cells using the Illumina MethylationEPIC Beadarray.

Results

We identified three latent profiles of women: a group with few risk factors (61%) and groups with elevated physical (26%) and psychological (13%) risk factors. Neonates born to women in higher risk subgroups had differential DNAm at 2 CpG sites. Higher cumulative prenatal risk was associated with methylation at 15 CpG sites, 12 of which were located in genes previously linked to physical and mental health and neurodevelopment.

Conclusion

We observed associations between prenatal risk factors and DNAm in very preterm infants using both person-centered and cumulative risk approaches. Epigenetics offers a potential biological indicator of prenatal risk exposure.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13148-021-01164-9.

Disentangling tumorigenesis-associated DNA methylation changes in colorectal tissues from those associated with ageing

Physiological ageing and tumorigenesis are both associated with epigenomic alterations in human tissue cells, the most extensively investigated of which entails de novo cytosine methylation (i.e., hypermethylation) within the CpG dinucleotides of CpG islands. Genomic regions that become hypermethylated during tumorigenesis are generally believed to overlap regions that acquire methylation in normal tissues as an effect of ageing. To define the extension of this overlap, we analysed the DNA methylomes of 48 large-bowel tissue samples taken from women of different ages during screening colonoscopy: 18 paired samples of normal and lesional tissues from donors harbouring a precancerous lesion and 12 samples of normal mucosa from tumour-free donors. Each sample was subjected to targeted, genome-wide bisulphite sequencing of ~2.5% of the genome, including all CpG islands. In terms of both its magnitude and extension along the chromatin, tumour-associated DNA hypermethylation in these regions was much more conspicuous than that observed in the normal mucosal samples from older (vs. younger) tumour-free donors. 83% of the ageing-associated hypermethylated regions (n = 2501) coincided with hypermethylated regions observed in tumour samples. However, 86% of the regions displaying hypermethylation in precancerous lesions (n = 16,772) showed no methylation changes in the ageing normal mucosa. The tumour-specificity of this latter hypermethylation was validated using published sets of data on DNA methylation in normal and neoplastic colon tissues. This extensive set of genomic regions displaying tumour-specific hypermethylation represents a rich vein of putative biomarkers for the early, non-invasive detection of colorectal tumours in women of all ages.

CUE: CpG impUtation ensemble for DNA methylation levels across the human methylation450 (HM450) and EPIC (HM850) BeadChip platforms

DNA methylation at CpG dinucleotides is one of the most extensively studied epigenetic marks. With technological advancements, geneticists can profile DNA methylation with multiple reliable approaches. However, profiling platforms can differ substantially in the CpGs they assess, consequently hindering integrated analysis across platforms. Here, we present CpG impUtation Ensemble (CUE), which leverages multiple classical statistical and modern machine learning methods, to impute from the Illumina HumanMethylation450 (HM450) BeadChip to the Illumina HumanMethylationEPIC (HM850) BeadChip. Data were analysed from two population cohorts with methylation measured both by HM450 and HM850: the Extremely Low Gestational Age Newborns (ELGAN) study (n = 127, placenta) and the VA Boston Posttraumatic Stress Disorder (PTSD) genetics repository (n = 144, whole blood). Cross-validation results show that CUE achieves the lowest predicted root-mean-square error (RMSE) (0.026 in PTSD) and the highest accuracy (99.97% in PTSD) compared with five individual methods tested, including k-nearest-neighbours, logistic regression, penalized functional regression, random forest, and XGBoost. Finally, among all 339,033 HM850-only CpG sites shared between ELGAN and PTSD, CUE successfully imputed 289,604 (85.4%) sites, where success was defined as RMSE < 0.05 and accuracy >95% in PTSD. In summary, CUE is a valuable tool for imputing CpG methylation from the HM450 to HM850 platform.

A deep embedded refined clustering approach for breast cancer distinction based on DNA methylation

Epigenetic alterations have an important role in the development of several types of cancer. Epigenetic studies generate a large amount of data, which makes it essential to develop novel models capable of dealing with large-scale data. In this work, we propose a deep embedded refined clustering method for breast cancer differentiation based on DNA methylation. In concrete, the deep learning system presented here uses the levels of CpG island methylation between 0 and 1. The proposed approach is composed of two main stages. The first stage consists in the dimensionality reduction of the methylation data based on an autoencoder. The second stage is a clustering algorithm based on the soft assignment of the latent space provided by the autoencoder. The whole method is optimized through a weighted loss function composed of two terms: reconstruction and classification terms. To the best of the authors’ knowledge, no previous studies have focused on the dimensionality reduction algorithms linked to classification trained end-to-end for DNA methylation analysis. The proposed method achieves an unsupervised clustering accuracy of 0.9927 and an error rate (%) of 0.73 on 137 breast tissue samples. After a second test of the deep-learning-based method using a different methylation database, an accuracy of 0.9343 and an error rate (%) of 6.57 on 45 breast tissue samples are obtained. Based on these results, the proposed algorithm outperforms other state-of-the-art methods evaluated under the same conditions for breast cancer classification based on DNA methylation data.

A machine learning approach to brain epigenetic analysis reveals kinases associated with Alzheimer's disease

Alzheimer's disease (AD) is influenced by both genetic and environmental factors; thus, brain epigenomic alterations may provide insights into AD pathogenesis. Multiple array-based Epigenome-Wide Association Studies (EWASs) have identified robust brain methylation changes in AD; however, array-based assays only test about 2% of all CpG sites in the genome. Here, we develop EWASplus, a computational method that uses a supervised machine learning strategy to extend EWAS coverage to the entire genome. Application to six AD-related traits predicts hundreds of new significant brain CpGs associated with AD, some of which are further validated experimentally. EWASplus also performs well on data collected from independent cohorts and different brain regions. Genes found near top EWASplus loci are enriched for kinases and for genes with evidence for physical interactions with known AD genes. In this work, we show that EWASplus implicates additional epigenetic loci for AD that are not found using array-based AD EWASs.

Genome-wide DNA methylation profiles provide insight into epigenetic regulation of red and white muscle development in Chinese perch Siniperca chuatsi

Fish skeletal muscles are composed of spatially well-separated fiber types, namely, red and white muscles with different physiological functions and metabolism. To compare the DNA methylation profiles of the two types of muscle tissues and identify potential candidate genes for the muscle growth and development under epigenetic regulation, genome-wide DNA methylation of the red and white muscle in Chinese perch Siniperca chuatsi were comparatively analyzed using bisulfate sequencing methods. An average of 0.9 billion 150-bp paired-end reads were obtained, of which 86% were uniquely mapped to the genome. Methylation mostly occurred at CG sites at a ratio of 94.43% in the red muscle and 93.16% in the white muscle. The mean methylation levels at C-sites were 5.95% in red muscle and 5.83% in white muscle, whereas the mean methylation levels of CG, CHG, and CHH were 73.23%, 0.62%, and 0.67% in red muscle, and 71.01%, 0.62%, and 0.67% in white muscle, respectively. A total of 4192 differentially methylated genes (DMGs) were identified significantly enriched in cell signaling pathways related to skeletal muscle differentiation and growth. Various muscle-related genes, including myosin gene isoforms and regulatory factors, are differentially methylated in the promoter region between the red and white muscles. Further analysis of the transcriptional expression of these genes showed that the muscle regulatory factors (myf5, myog, pax3, pax7, and twitst2) and myosin genes (myh10, myh16, myo18a, myo7a, myo9a, and myl3) were differentially expressed between the two kinds of muscles, consistent with the DNA methylation analysis results. ELISA assays confirmed that the level of 5mC in red muscle was significantly higher than in white muscle (P < 0.05). The RT-qPCR assays revealed that the expression levels of the three DNA methylation transferase (dnmt) subtypes, dnmt1, dnmt3ab, and dnmt3bb1, were significantly higher in red muscle than in white muscle. The higher DNA methylation levels in the red muscle may result from higher DNA methylation transferase expression in the red muscles. Thus, this study might provide a theoretical foundation to better understand epigenetic regulation in the growth and development of red and white muscles in animals, at least in Chinese perch fish.

Determining subpopulation methylation profiles from bisulfite sequencing data of heterogeneous samples using DXM

Epigenetic changes, such as aberrant DNA methylation, contribute to cancer clonal expansion and disease progression. However, identifying subpopulation-level changes in a heterogeneous sample remains challenging. Thus, we have developed a computational approach, DXM, to deconvolve the methylation profiles of major allelic subpopulations from the bisulfite sequencing data of a heterogeneous sample. DXM does not require prior knowledge of the number of subpopulations or types of cells to expect. We benchmark DXM's performance and demonstrate improvement over existing methods. We further experimentally validate DXM predicted allelic subpopulation-methylation profiles in four Diffuse Large B-Cell Lymphomas (DLBCLs). Lastly, as proof-of-concept, we apply DXM to a cohort of 31 DLBCLs and relate allelic subpopulation methylation profiles to relapse. We thus demonstrate that DXM can robustly find allelic subpopulation methylation profiles that may contribute to disease progression using bisulfite sequencing data of any heterogeneous sample.

mTORC1 stimulates cell growth through SAM synthesis and m6A mRNA-dependent control of protein synthesis

The mechanistic target of rapamycin complex 1 (mTORC1) regulates metabolism and cell growth in response to nutrient, growth, and oncogenic signals. We found that mTORC1 stimulates the synthesis of the major methyl donor, S-adenosylmethionine (SAM), through the control of methionine adenosyltransferase 2 alpha (MAT2A) expression. The transcription factor c-MYC, downstream of mTORC1, directly binds to intron 1 of MAT2A and promotes its expression. Furthermore, mTORC1 increases the protein abundance of Wilms' tumor 1-associating protein (WTAP), the positive regulatory subunit of the human N6-methyladenosine (m6A) RNA methyltransferase complex. Through the control of MAT2A and WTAP levels, mTORC1 signaling stimulates m6A RNA modification to promote protein synthesis and cell growth. A decline in intracellular SAM levels upon MAT2A inhibition decreases m6A RNA modification, protein synthesis rate, and tumor growth. Thus, mTORC1 adjusts m6A RNA modification through the control of SAM and WTAP levels to prime the translation machinery for anabolic cell growth.

An efficient ensemble method for missing value imputation in microarray gene expression data

BACKGROUND

The genomics data analysis has been widely used to study disease genes and drug targets. However, the existence of missing values in genomics datasets poses a significant problem, which severely hinders the use of genomics data. Current imputation methods based on a single learner often explores less known genomic data information for imputation and thus causes the imputation performance loss.

RESULTS

In this study, multiple single imputation methods are combined into an imputation method by ensemble learning. In the ensemble method, the bootstrap sampling is applied for predictions of missing values by each component method, and these predictions are weighted and summed to produce the final prediction. The optimal weights are learned from known gene data in the sense of minimizing a cost function about the imputation error. And the expression of the optimal weights is derived in closed form. Additionally, the performance of the ensemble method is analytically investigated, in terms of the sum of squared regression errors. The proposed method is simulated on several typical genomic datasets and compared with the state-of-the-art imputation methods at different noise levels, sample sizes and data missing rates. Experimental results show that the proposed method achieves the improved imputation performance in terms of the imputation accuracy, robustness and generalization.

CONCLUSION

The ensemble method possesses the superior imputation performance since it can make use of known data information more efficiently for missing data imputation by integrating diverse imputation methods and learning the integration weights in a data-driven way.

DNA Modification Patterns within the Transposable Elements of the Fig (Ficus carica L.) Genome

Transposable element activity can be harmful to the host’s genome integrity, but it can also provide selective advantages. One strategy to cope with transposons is epigenetic control through DNA base modifications. We report the non-canonic DNA modification dynamics of fig (Ficus carica L.) by exploiting high-quality genome reference and related N4-methylcytosine (4mC) and N6-methyladenine (6mA) data. Overall, 1.49% of transposon nucleotides showed either 4mC or 6mA modifications: the 4mC/6mA ratio was similar in Class I and Class II transposons, with a prevalence of 4mC, which is comparable to coding genes. Different percentages of 4mC or 6mA were observed among LTR-retrotransposon lineages and sub-lineages. Furthermore, both the Copia and Gypsy retroelements showed higher modification rates in the LTR and coding regions compared with their neighbour regions. Finally, the unconventional methylation of retrotransposons is unrelated to the number of close genes, suggesting that the 4mC and 6mA frequency in LTR-retrotransposons should not be related to transcriptional repression in the adjacency of the element. In conclusion, this study highlighted unconventional DNA modification patterns in fig transposable elements. Further investigations will focus on functional implications, in regards to how modified retroelements affect the expression of neighbouring genes, and whether these epigenetic markers can spread from repeats to genes, shaping the plant phenotype.