Biology subjects

Markov, Y.

Publications and source records attributed to Markov, Y..

7 recordsLinked to original sources

bioRxiv2025

BrainBridge Characterizes Key Factors affecting Alzheimer's Disease and Associated Phenotypes

Single-cell RNA sequencing (scRNA-seq) has significantly advanced our understanding of Alzheimers disease and aging by revealing cellular heterogeneity and shifts in cell-type composition between diseased/old and healthy/young individuals. However, few existing studies utilize the rich information in single-cell transcriptomic atlases for robust patient-level modeling and biological feature selection. To address this gap, we present BrainBridge, a deep learning-based framework designed to integrate atlas-scale single-cell transcriptomic data with phenotypic information to model the biomolecular complexity of the human brain. BrainBridge functions both as a powerful predictor and an embedding model for representing sample-level expression profiles and covariates through comprehensive benchmarking. We also demonstrate its effectiveness in prioritizing key genes and cell types associated with disease progression, aging, and sex differences. We further validate our findings using external resources, including genome-wide and epigenome-wide association studies (GWAS and EWAS), spatial transcriptomics, and perturb-seq experiments. Finally, we deploy BrainBridge within an interactive, agent-powered interface that enables intuitive and user-friendly model interactions, promoting broader accessibility and application in biomedical research.

bioinformatics↗

bioRxiv2025

TranslAGE: A Comprehensive Platform for Systematic Validation of Epigenetic Aging Biomarkers

Epigenetic clocks are powerful biomarkers of biological aging, however, their performance varies across studies and contexts. Current limitations include siloed datasets, inconsistent validation methods, and the absence of a standardized framework for systematic comparison. Here, we introduce TranslAGE: a publicly available online resource that addresses this gap by harmonizing 179 human blood DNA methylation datasets and precalculating a suite of 41 epigenetic biomarker scores for each of the >42,000 total samples. Users can explore these data through interactive dashboards that evaluate four fundamental performance domains: Stability, Treatment response, Associations, and Risk, collectively forming the STAR framework. Stability quantifies robustness to multiple types of technical and biological noise. Treatment response measures biomarker sensitivity to aging interventions and environmental exposures. Associations capture cross-sectional relationships with age, demographics, disease, and other phenotypes, and Risk assesses predictive power for future functional decline, morbidity and mortality. The STAR framework unifies these test metrics into a single composite scoring system that enables researchers to identify, benchmark, and validate biomarkers best suited to their scientific or clinical applications. TranslAGE will be continually updated, with rapid scaling by adding datasets, biomarkers, or analyses. By providing harmonized datasets, precomputed biomarker scores, and interactive data tools, TranslAGE establishes the first standardized, reproducible framework for benchmarking epigenetic aging biomarkers across populations, and accelerates the translation toward clinical use.

genomics↗

bioRxiv2025

Beyond the Genotype: A Multi-Omic Analysis of APOEe4's Role in Alzheimer's Disease

Alzheimers disease (AD) is characterized by widespread molecular dysregulation, with the APOEe4 allele recognized as its strongest genetic risk factor. However, the mechanisms by which APOEe4 drives distinct molecular changes - whether by exacerbating pathology or triggering compensatory responses - remain incompletely understood. We generated and analyzed proteomic, epigenetic, and genetic data from post-mortem dorsolateral prefrontal cortex samples of a uniquely APOEe4-enriched subset of the Religious Orders Study and Memory and Aging Project (ROSMAP). Specifically, we generated DIA LC-MS proteomic data (n = 302), analyzed previously generated DNA methylation profiles from our group (n = 310), and used published whole-genome sequencing data (n = 254) to compute polygenic risk scores (PRS). In this cohort, 69% (n = 214) were APOEe4 carriers, and 19.6% (n = 42) of them showed no pathological evidence of AD based on NIA-Reagan criteria, enabling identification of APOEe4-related risk and resilience mechanisms. In the absence of AD, APOEe4 carriers exhibited lower levels of 27 proteins, suggesting early synaptic (e.g., VAMP1, SYN3, CASKIN1) and metabolic (e.g., GLUD1, PI4KA) vulnerability. By contrast, APOEe4 carriers with AD displayed marked upregulation of inflammatory and proteostatic proteins (e.g., GNAO1, AHNAK, FGG, HEBP1, APEX1, RAB4A, SLC12A5, LRP1, BAG6) and hypermethylation of cg06329447 in ELAVL4. Network analyses highlighted convergent disruptions in synaptic transmission, metabolism, and proteostasis - key pathways altered in APOEe4-associated AD. Mediation analyses identified GRIPAP1 and GSTK1 as top protein mediators (accounting for [~]26-33% of APOEe4s effect), with VAMP1, CASKIN1, DPP3, SYN3, and FGG each contributing [~]9-15%. ELAVL4 hypermethylation also mediated [~]12% of the APOEe4 effect, linking epigenetic dysregulation to disease risk. To assess whether the identified proteins reflected broader genetic risk for AD or were specific to APOEe4, we calculated PRS both excluding and including the APOE genomic region. While the non-APOE PRS showed no association with identified molecular markers, the APOE-inclusive PRS was significantly associated with eight AD-related proteins in carriers, indicating they are not explained by polygenic risk outside of APOE. Finally, predictive modeling stratified by APOEe4 status revealed that in non-carriers, PRS most effectively classified AD (AUC = 0.73), whereas in carriers, proteomic and epigenetic markers outperformed PRS (AUC up to 0.74). Together, these findings demonstrate that APOEe4 confers AD risk through early synaptic and metabolic disruptions and later-stage inflammatory and epigenetic changes, laying the groundwork for genotype-tailored biomarker development and therapeutic strategies. VISUAL ABSTRACT O_FIG O_LINKSMALLFIG WIDTH=156 HEIGHT=200 SRC="FIGDIR/small/682426v1_ufig1.gif" ALT="Figure 1"> View larger version (31K): org.highwire.dtl.DTLVardef@1477a2forg.highwire.dtl.DTLVardef@1a6d173org.highwire.dtl.DTLVardef@1010821org.highwire.dtl.DTLVardef@bcde3a_HPS_FORMAT_FIGEXP M_FIG C_FIG

pathology↗

bioRxiv2025

Biological versus Technical Reliability of Epigenetic Clocks and Implications for Disease Prognosis and Intervention Response

DNA methylation-based aging biomarkers, or epigenetic clocks, are increasingly used to estimate biological age and predict health outcomes. Their translational utility, however, depends not only on predictive accuracy but also on reliability, the ability to provide consistent results across technical replicates and repeated biological measures. Here, we leveraged the TranslAGE platform to comprehensively evaluate the technical and biological reliability of 18 Epigenetic clocks, including chronological predictors, mortality predictors, pace-of-aging measures, reliable variants, and newer explainable clocks. Technical reliability was quantified across four independent datasets. For standard replicate assays on EPIC and 450K arrays, nearly all clocks achieved excellent technical reproducibility. However, some clocks showed dramatic drops in technical reliability based on differences in slide position and DNA extraction protocol. PC-based clocks, especially PCGrimAge and SystemsAge remained technically reliable in all cases. In contrast, biological reliability, measured across repeated samples collected within hours, before and after meals, under acute stress, across environmental exposures, and over days, was markedly lower, with most clocks showing only moderate stability. PCGrimAge was the only clock with good ICC > 0.75 for biological reliability. Importantly, technical reproducibility did not predict biological reliability; clocks that were technically robust often proved biologically unreliable. We further demonstrated that reliability directly constrains downstream applications. Clocks with higher ICCs produced more stable prognostic associations with cognitive decline and more consistent responsiveness to a vegan diet intervention, whereas unreliable clocks yielded highly variable or spurious effects. Together, these findings reveal that technical reliability is not enough: biological reliability remains a critical limitation for many DNA methylation clocks that constrains their utility, and our work provides a roadmap for prioritizing next-generation clocks most suited for clinical translation.

bioinformatics↗

bioRxiv2024

DNAm aging biomarkers are responsive: Insights from 51 longevity interventional studies in humans

Aging biomarkers can potentially allow researchers to rapidly monitor the impact of an aging intervention, without the need for decade-spanning trials, by acting as surrogate endpoints. Prior to testing whether aging biomarkers may be useful as surrogate endpoints, it is first necessary to determine whether they are responsive to interventions that target aging. Epigenetic clocks are aging biomarkers based on DNA methylation with prognostic value for many aging outcomes. Many individual studies are beginning to explore whether epigenetic clocks are responsive to interventions. However, the diversity of both interventions and epigenetic clocks in different studies make them difficult to compare systematically. Here, we curate TranslAGE-Response, a harmonized database of 51 public and private longitudinal interventional studies and calculate a consistent set of 16 prominent epigenetic clocks for each study, along with 95 other DNAm biomarkers that help explain changes in each clock. With this database, we discover patterns of responsiveness across a variety of interventions and DNAm biomarkers. For example, clocks trained to predict mortality or pace of aging have the strongest response across all interventions and show consistent agreement with each other, pharmacological and lifestyle interventions drive the strongest response from DNAm biomarkers, and study population and study duration are key factors in driving responsiveness of DNAm biomarkers in an intervention. Some classes of interventions such as TNF-alpha inhibitors have strong, consistent effects across multiple studies, while others such as senolytic drugs have inconsistent effects. Clocks with multiple sub-scores (i.e. "explainable clocks") provide specificity and greater mechanistic insight into responsiveness of interventions than single-score clocks. Our work can help the geroscience field design future clinical trials, by guiding the choice of interventions, specific subsets of epigenetic clocks to minimize multiple testing, study duration, study population, and sample size, with the eventual aim of determining whether epigenetic clocks can be used as surrogate endpoints.

systems biology↗

bioRxiv2023

Stochastic Epigenetic Mutations: Reliable Detection and Cardiovascular Associations

Stochastic Epigenetic Mutations (SEMs) have been proposed as novel aging biomarkers that have the potential to capture heterogeneity in age-related DNA methylation (DNAme) changes. SEMs are defined as outlier methylation patterns at cytosine-guanine dinucleotide (CpG) sites, categorized as hypermethylated (hyperSEM) or hypomethylated (hypoSEM) relative to a reference. While individual SEMs are rarely consistent across subjects, the SEM load - the total number of SEMs - increases with age. However, given poor technical reliability of measurement for many DNA methylation sites, we posited that many outliers might represent technical noise. Our study of whole blood samples from 36 individuals, each measured twice, found that 23.3% of hypoSEM and 45.6% hyperSEM are not shared between replicates. This diminishes the reliability of SEM loads, where intraclass correlation coefficients are 0.96 for hypoSEM and 0.90 for hyperSEM. We linked SEM reliability to multiple factors, including blood cell type composition, probe beta-value statistics, and presence of SNPs. A machine learning approach, leveraging these factors, filtered unreliable SEMs, enhancing reliability in a separate dataset of technical replicates from 128 individuals. Analysis of the Framingham Heart Study confirmed previously reported SEM association with mortality and revealed novel connections to cardiovascular disease. We discover that associations with aging outcomes are primarily driven by hypoSEMs at baseline methylated probes and hyperSEMs at baseline unmethylated probes, which are the same subsets that demonstrate highest technical reliability. These aging associations are preserved after filtering out unreliable SEMs and are enhanced after adjusting for blood cell composition. Finally, we utilize these insights to formulate best practices for SEM detection and introduce a novel R package, SEMdetectR, which utilizes parallel programming for efficient SEM detection with comprehensive options for detection, filtering, and analysis.

bioinformatics↗

bioRxiv2017

Genomic analysis of human polymorphisms affecting drug-protein interactions

Human genetic variability is thought to account for a substantial fraction of individual biochemical characteristics - in biomedical sense, of individual drug response. However, only a handful of human genetic variants have been linked to medication outcomes. Here, we combine data on drug-protein interactions and human genome sequences to assess the impact of human variation on their binding affinity. Using data from the complexes of FDA-drugs and drug-like compounds, we predict SNPs substantially affecting the protein-ligand binding affinities. We estimate that an average individual carries ~6 SNPs affecting ~5 different FDA-approved drugs from among all of the approved compounds. SNPs affecting drug-protein binding affinity have low frequency in the population indicating that the genetic component for many ADEs may be highly personalized with each individual carrying a unique set of relevant SNPs. The reduction of ADEs, therefore, may primarily rely on the application of computational genome analysis in the clinic rather than the experimental study of common SNPs.

bioinformatics↗