OSCR

Epigenetic Aging in Brain Tissue of the Self-Fertilizing Vertebrate, <i>Kryptolebias marmoratus</i>.

Code ↔ Paper

7 matches between paragraphs of the paper and lines of its authors' code, computed by the harvester (lexical-v1). Click a colored paragraph or line to see its counterpart.

The 7 matches · 1 of them tie a paragraph to a whole file, not to given lines: a weak match, whose lines are not tinted
  1. [1] § Material and Methods › Bioinformatics › Methylation Analysis ↔ Process_clock.R, the whole file · a weak match · score 0.76 · processBismarkAln, read.context, ASM164957v2, methylKit, assembly, Bam
  2. [2] § Material and Methods › Bioinformatics › Methylation Analysis ↔ pre-clock-young.R, lines 61–107 · score 0.68 · processBismarkAln, read.context, ASM164957v2, assembly, Bam
  3. [3] § Material and Methods › Bioinformatics › Predicting Age From CpG Methylation ↔ Epigenetic_clock.R, lines 488–559 · score 0.58 · Absolute Error, epigenetic age, hyperparameters, clock, MAE, chronological
  4. [4] § Material and Methods › Bioinformatics › Predicting Age From CpG Methylation ↔ clock-young-boucle-file.R, lines 408–490 · score 0.56 · Absolute Error, cross validation, clock, MAE, chronological, epigenetic
  5. [5] § Material and Methods › Bioinformatics › Predicting Age From CpG Methylation ↔ Epigenetic_clock.R, lines 1–71 · score 0.53 · percMethylation, methylKit, CpGs, seeds, age
  6. [6] § Material and Methods › Bioinformatics › Predicting Age From CpG Methylation ↔ clock-young-boucle-file.R, lines 1–68 · score 0.51 · percMethylation, methylKit, seeds, age
  7. [7] § Results › Age‐Associated CpG Sites Identified by RRBS ↔ clock-young-boucle-file.R, lines 408–490 · score 0.50 · absolute error, Cross Validation, LOOCV, MAE, training, model

Paper

Loaded from Europe PMC by your browser, not stored by OSCR: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The authors' code

R · 494 lines · 16 KB · no license · 3 matches

  1. ### R SetUp
  3. options(expressions = 5e5)
  5. cran_mirror <- Sys.getenv("R_CRAN_MIRROR", "http://cran.rstudio.com/")
  6. options(repos = cran_mirror)
  8. if (!require("BiocManager", quietly = TRUE))
  9. install.packages("BiocManager")
  11. BiocManager::install("methylKit")
  12. BiocManager::install("Metrics")
  13. BiocManager::install("ggpubr")
  14. BiocManager::install("doParallel")
  16. library(BiocManager)
  17. library(stringi)
  18. library(pacman)
  19. library(methylKit)
  20. library(ade4)
  21. library(FactoMineR)
  22. library(devtools)
  23. library(factoextra)
  24. library(tibble)
  25. library(caret)
  26. library(tibble)
  27. library(Metrics)
  28. library(ggpubr)
  29. library(stats)
  30. library(ggplot2)
  31. library(openxlsx)
  33. set.seed(935)
  35. ### Loading Data
  37. myobj <- readRDS("Clock-new-young.RDS")
  39. filtered.myobj <- filterByCoverage(myobj,lo.count=15,lo.perc= NULL, hi.count=NULL,hi.perc=99.9)
  40. meth <- methylKit::unite(filtered.myobj, destrand=T, min.per.group = NULL)
  41. check1 <- head(meth)
  43. write.csv(meth, "meth.csv")
  44. write.csv(check1, "check1.csv")
  46. pm=percMethylation(meth)
  47. check2 <- pm
  48. write.csv(check2, "check2.csv")
  50. pm_order <- t(pm)
  51. pm_order <- data.frame(pm_order)
  52. pm_ID <- rownames_to_column(pm_order, var = "ID")
  53. pm_meth <- pm_ID[,-1]
  55. Age <- c(869, 811, 811, 811, 748, 749, 749, 749, 749, 683, 683, 683, 683, 683, 683, 786, 786, 882, 882, 513, 510, 510, 570, 450, 450, 566, 450, 449, 574, 338, 570, 390, 390, 390, 338, 338, 390, 331, 331, 331, 331, 270, 270, 270, 270, 211, 211, 212, 211, 211, 211, 151, 151, 150, 150, 151, 151, 91, 90, 90, 90, 90, 89, 63, 63, 63, 63, 63, 62, 121, 120, 120, 120, 120, 120, 120)
  57. pm_round <- round(pm_meth, digits = 0)
  59. pm_complete <- cbind(Age, pm_round)
  60. pm_complete$Age <- log(pm_complete$Age)
  62. rm(pm)
  63. rm(pm_order)
  64. rm(pm_ID)
  65. rm(filtered.myobj)
  66. rm(meth)
  68. gc() # empty the environment of the rm object to save space
  70. ### Check for NA
  72. data <- na.omit(pm_complete)
  73. age <- data[, "Age"]
  74. cpGs <- data[, 3:ncol(data)]
  76. ### Correlation Age ~ CpG
  78. results <- data.frame(CpG = colnames(cpGs), correlation = NA, pval = NA)
  80. for (i in 1:ncol(cpGs)) {
  81. corr_test <- cor.test(age, cpGs[, i])
  82. results$correlation[i] <- corr_test$estimate
  83. results$pval[i] <- corr_test$p.value
  84. }
  86. results$adj_pval <- p.adjust(results$pval, method = "BH")
  88. significant_cpGs <- results[results$adj_pval < 0.05, ]
  89. write.csv(significant_cpGs, "CpG_significatifs.csv", row.names = FALSE)
  91. common_cpgs <- intersect(colnames(pm_complete), significant_cpGs$CpG)
  93. significant_cpG_data <- pm_complete[, common_cpgs, drop = FALSE]
  94. significant_cpG_data$Age <- cbind(Age, significant_cpG_data)
  95. significant_cpG_data$Age <- log(Age)
  97. ### Elastic Net Regression (first step using all CpG)
  99. which_training <- createDataPartition(significant_cpG_data$Age, p = 0.80)[[1]]
  100. training_data <- significant_cpG_data[which_training,]
  101. testing_data <- significant_cpG_data[-which_training,]
  103. ctrl <- trainControl(method = "cv", number = 10)
  105. fitted_model <- train(
  106. Age ~ .,
  107. data = training_data,
  108. method = "glmnet",
  109. trControl = ctrl,
  110. preProc = c("center", "scale", "nzv")
  111. )
  113. fitted_model
  115. model_predictions <- predict(fitted_model, testing_data)
  116. write.csv(model_predictions, "model_predictions.csv")
  118. model_predictions_tr <- predict(fitted_model, training_data)
  119. write.csv(model_predictions_tr, "model_predictions_training.csv")
  121. variable_importance <- varImp(fitted_model)$importance$Overall
  122. importance_cutoff <- 1
  123. key_variables <- rownames(varImp(fitted_model)$importance)[varImp(fitted_model)$importance > importance_cutoff]
  124. key_importances <- variable_importance[variable_importance > importance_cutoff]
  126. if (length(key_variables) != length(key_importances)) {
  127. stop("Length key_variables and key_importances are different")
  128. }
  130. key_variables_with_importance <- data.frame(
  131. Variable = key_variables,
  132. Importance = key_importances
  133. )
  135. write.csv(key_variables_with_importance, "key_variables_with_importance.csv")
  137. ### Take the methylation level that correspond to the CpG with the respective importance cutoff
  139. missing_cpgs <- setdiff(key_variables, colnames(pm_complete))
  140. if (length(missing_cpgs) > 0) {
  141. cat("Les CpG suivants ne sont pas présents dans pm_complete :", paste(missing_cpgs, collapse = ", "), "\n")
  142. }
  144. variables_to_select <- c("Age", key_variables)
  145. selected_columns <- pm_complete[, variables_to_select, drop = FALSE]
  147. ### List of CpG in selected_columns, excluding Age
  148. cpg_columns <- colnames(selected_columns)[-1]
  150. RMSE_te <- data.frame(RMSE = RMSE(model_predictions, testing_data$Age),
  151. Rsquare = R2(model_predictions, testing_data$Age),
  152. MAE = mae(testing_data$Age, model_predictions),
  153. PEARSON = cor(testing_data$Age, model_predictions, method = 'pearson'))
  155. write.csv(RMSE_te, "RMSE_te.csv")
  157. RMSE_tr <- data.frame(RMSE = RMSE(model_predictions_tr, training_data$Age),
  158. Rsquare = R2(model_predictions_tr, training_data$Age),
  159. MAE = mae(training_data$Age, model_predictions_tr),
  160. PEARSON = cor(training_data$Age, model_predictions_tr, method = 'pearson'))
  162. write.csv(RMSE_tr, "RMSE_tr.csv")
  164. ### First loop: importance cutoff from 5 to 75, by 5 (very general)
  165. # cutoff_values <- seq(5, 75, by = 5)
  166. # Exact same code as below. Allows us to determine the a smaller cutoff interval, worth looking at in details
  168. ### Second loop: importance cutoff from 25 to 40, by 1 (more precise)
  169. cutoff_values <- seq(25, 40, by = 1)
  171. # Initialise list to stock results
  172. rmse_te_list <- list()
  173. rmse_tr_list <- list()
  174. predictions_te_list <- list()
  175. predictions_tr_list <- list()
  176. selected_columns_list <- list()
  178. # Add missing columns to data frame
  179. align_columns <- function(df_list) {
  180. all_columns <- unique(unlist(lapply(df_list, colnames)))
  181. lapply(df_list, function(df) {
  182. missing_cols <- setdiff(all_columns, colnames(df))
  183. for (col in missing_cols) {
  184. df[[col]] <- NA
  185. }
  186. df <- df[, all_columns]
  187. return(df)
  188. })
  189. }
  191. cat("Loop begins...\n")
  193. for (cutoff in cutoff_values) {
  195. # Step 1: importance cutoff
  196. variable_importance <- varImp(fitted_model)$importance$Overall
  197. key_variables <- rownames(varImp(fitted_model)$importance)[varImp(fitted_model)$importance > cutoff]
  198. key_importances <- variable_importance[variable_importance > cutoff]
  200. # Step 2: Select column for the subset
  201. variables_to_select <- c("Age", key_variables)
  202. subset_pm <- pm_complete[, variables_to_select, drop = FALSE]
  204. # Add selected column to the list
  205. selected_columns_list[[as.character(cutoff)]] <- cbind(Cutoff = cutoff, subset_pm)
  207. # Step 3: Divide into training and testing
  208. try({
  209. subset_training_data <- subset_pm[which_training,]
  210. subset_testing_data <- subset_pm[-which_training,]
  211. cat("Étape 3 réussie : Division des données.\n")
  212. }, silent = FALSE)
  214. # Step 4: Model training
  215. subset_fitted_model <- train(
  216. Age ~ .,
  217. data = subset_training_data,
  218. method = "glmnet",
  219. trControl = ctrl
  220. )
  222. # Step 5: Model testing (prediction)
  223. subset_model_predictions_te <- predict(subset_fitted_model, testing_data)
  224. subset_model_predictions_tr <- predict(subset_fitted_model, training_data)
  226. # Add predictions to list
  227. predictions_te_list[[as.character(cutoff)]] <- data.frame(
  228. Cutoff = cutoff,
  229. Actual = testing_data$Age,
  230. Predicted = subset_model_predictions_te
  231. )
  233. predictions_tr_list[[as.character(cutoff)]] <- data.frame(
  234. Cutoff = cutoff,
  235. Actual = training_data$Age,
  236. Predicted = subset_model_predictions_tr
  237. )
  239. # Step 6: Calculate model parameters
  240. RMSE_subset_te <- data.frame(
  241. Cutoff = cutoff,
  242. RMSE = RMSE(subset_model_predictions_te, testing_data$Age),
  243. Rsquare = R2(subset_model_predictions_te, testing_data$Age),
  244. MAE = mae(testing_data$Age, subset_model_predictions_te),
  245. PEARSON = cor(testing_data$Age, subset_model_predictions_te, method = 'pearson')
  246. )
  248. RMSE_subset_tr <- data.frame(
  249. Cutoff = cutoff,
  250. RMSE = RMSE(subset_model_predictions_tr, training_data$Age),
  251. Rsquare = R2(subset_model_predictions_tr, training_data$Age),
  252. MAE = mae(training_data$Age, subset_model_predictions_tr),
  253. PEARSON = cor(training_data$Age, subset_model_predictions_tr, method = 'pearson')
  254. )
  256. # Add them to the list
  257. rmse_te_list[[as.character(cutoff)]] <- RMSE_subset_te
  258. rmse_tr_list[[as.character(cutoff)]] <- RMSE_subset_tr
  259. }
  261. cat("End of loop...\n")
  263. # Align columns before combining
  264. all_rmse_te <- do.call(rbind, rmse_te_list)
  265. all_rmse_tr <- do.call(rbind, rmse_tr_list)
  267. all_predictions_te <- do.call(rbind, align_columns(predictions_te_list))
  268. all_predictions_tr <- do.call(rbind, align_columns(predictions_tr_list))
  269. all_selected_columns <- do.call(rbind, align_columns(selected_columns_list))
  271. # Save into a single Excel file
  272. write.xlsx(list(
  273. "Test_Set_Metrics" = all_rmse_te,
  274. "Training_Set_Metrics" = all_rmse_tr,
  275. "Test_Set_Predictions" = all_predictions_te,
  276. "Training_Set_Predictions" = all_predictions_tr,
  277. "Selected_Columns" = all_selected_columns
  278. ), file = "Consolidated_Model_Results.xlsx")
  280. cat("End with success.\n")
  282. #### Final model, with the optimized cutoff
  283. # Add graph and figure
  285. # Step 1
  286. cutoff <- 31
  288. variable_importance <- varImp(fitted_model)$importance$Overall
  289. key_variables <- rownames(varImp(fitted_model)$importance)[varImp(fitted_model)$importance > cutoff]
  290. key_importances <- variable_importance[variable_importance > cutoff]
  292. key_variables_with_importance <- data.frame(
  293. Variable = key_variables,
  294. Importance = key_importances
  295. )
  297. write.csv(key_variables_with_importance, paste0("key_variables_with_importance_final.csv"))
  299. # Step 2
  300. variables_to_select <- c("Age", key_variables)
  301. subset_pm <- pm_complete[, variables_to_select, drop = FALSE]
  303. write.csv(subset_pm, paste0("selected_columns_final.csv"), row.names = FALSE)
  305. # Step 3
  306. subset_training_data <- subset_pm[which_training,]
  307. subset_testing_data <- subset_pm[-which_training,]
  309. # Step 4
  310. subset_fitted_model <- train(
  311. Age ~ .,
  312. data = subset_training_data,
  313. method = "glmnet",
  314. trControl = ctrl
  315. )
  317. # Step 5
  318. subset_model_predictions_te <- predict(subset_fitted_model, testing_data)
  319. write.csv(subset_model_predictions_te, paste0("subset_model_predictions_final.csv"))
  321. subset_model_predictions_tr <- predict(subset_fitted_model, training_data)
  322. write.csv(subset_model_predictions_tr, paste0("subset_model_predictions_final_training.csv"))
  324. # Step 6
  325. RMSE_subset_te <- data.frame(
  326. RMSE = RMSE(subset_model_predictions_te, testing_data$Age),
  327. Rsquare = R2(subset_model_predictions_te, testing_data$Age),
  328. MAE = mae(testing_data$Age, subset_model_predictions_te),
  329. PEARSON = cor(testing_data$Age, subset_model_predictions_te, method = 'pearson')
  330. )
  331. write.csv(RMSE_subset_te, paste0("RMSE_subset_final_te.csv"))
  333. RMSE_subset_tr <- data.frame(
  334. RMSE = RMSE(subset_model_predictions_tr, training_data$Age),
  335. Rsquare = R2(subset_model_predictions_tr, training_data$Age),
  336. MAE = mae(training_data$Age, subset_model_predictions_tr),
  337. PEARSON = cor(training_data$Age, subset_model_predictions_tr, method = 'pearson')
  338. )
  339. write.csv(RMSE_subset_tr, paste0("RMSE_subset_final_tr.csv"))
  341. # Step 7: PCA
  343. pdf("predictions_ggplot_young_log_final.pdf", width = 5, height = 5)
  344. ggplot() +
  345. geom_point(
  346. aes(x = testing_data$Age, y = model_predictions),
  347. color = "red", size = 2
  348. ) +
  349. geom_point(
  350. aes(x = training_data$Age, y = predict(fitted_model)),
  351. color = "black", alpha = 0.2, size = 2
  352. ) +
  353. geom_abline(intercept = 0, slope = 1, linetype = "dashed") +
  354. labs(
  355. title = "Model Predictions",
  356. x = "log(Actual Age)",
  357. y = "log(Predicted Age)"
  358. ) +
  359. xlim(range(c(pm_complete$Age, model_predictions))) +
  360. ylim(range(c(pm_complete$Age, model_predictions))) +
  361. theme_minimal()
  362. dev.off()
  364. ###
  366. data_pca <- subset_pm[, -1] # Remove Age
  367. acp <- dudi.pca(data_pca, scannf = FALSE, nf = 3)
  369. pdf("scatter_acp_final.pdf", width = 5, height = 5)
  370. scatter(acp)
  371. dev.off()
  373. pdf("corcircle_acp_final.pdf", width = 5, height = 5)
  374. s.corcircle(acp$co, xax = 1, yax = 2, sub = paste0("Correlation according to plan 1-2, Cutoff = ", cutoff))
  375. dev.off()
  377. res.pca <- PCA(data_pca, graph = FALSE)
  378. pdf("fviz_pca_ind_log_final.pdf", width = 5, height = 5)
  379. fviz_pca_ind(res.pca, geom.ind = c("point"),
  380. col.ind = subset_pm$Age,
  381. gradient.cols = c("purple4", "orange"),
  382. legend.title = "Age")
  383. dev.off()
  385. # Step 8: Prediction visualisation
  387. pdf("predictions_ggplot_final.pdf", width = 5, height = 5)
  388. ggplot() +
  389. geom_point(
  390. aes(x = testing_data$Age, y = model_predictions),
  391. color = "red", size = 2
  392. ) +
  393. geom_point(
  394. aes(x = training_data$Age, y = predict(fitted_model)),
  395. color = "black", alpha = 0.2, size = 2
  396. ) +
  397. geom_abline(intercept = 0, slope = 1, linetype = "dashed") +
  398. labs(
  399. title = "Model Predictions",
  400. x = "log(Actual Age)",
  401. y = "log(Predicted Age)"
  402. ) +
  403. xlim(range(c(pm_complete$Age, model_predictions))) +
  404. ylim(range(c(pm_complete$Age, model_predictions))) +
  405. theme_minimal()
  406. dev.off()
  408. # Step 9: Visualisation of selected CpG
  410. cpg_columns <- colnames(subset_pm)[-1]
  412. pdf("CpG_graphs_final.pdf", width = 5, height = 5)
  414. for (cpg in cpg_columns) {
  415. model <- lm(selected_columns[[cpg]] ~ selected_columns$Age)
  416. r_squared <- summary(model)$r.squared
  417. plot <- ggplot(selected_columns, aes(x = Age, y = .data[[cpg]])) +
  418. geom_point() +
  419. geom_smooth(method = "lm") +
  420. labs(title = paste("Relation between Age and", cpg),
  421. x = "log(Chronological Age)", y = "Methylation level") +
  422. annotate("text", x = max(selected_columns$Age, na.rm = TRUE),
  423. y = max(selected_columns[[cpg]], na.rm = TRUE),
  424. label = paste("R² =", round(r_squared, 3)),
  425. hjust = 1.1, vjust = 1.1, size = 4, color = "blue") +
  426. theme_minimal()
  428. print(plot)
  429. }
  431. # Step 10 : Leave-One-Out-Cross-Validation
  433. ctrl <- trainControl(method = "LOOCV")
  435. fitted_model_loocv <- train(
  436. Age ~ .,
  437. data = training_data,
  438. method = "glmnet",
  439. trControl = ctrl,
  440. preProc = c("center", "scale", "nzv")
  441. )
  443. model_predictions_loocv_te <- predict(fitted_model_loocv, testing_data)
  444. write.csv(model_predictions_loocv_te, "model_predictions_loocv_te.csv")
  446. model_predictions_loocv_tr <- predict(fitted_model_loocv, training_data)
  447. write.csv(model_predictions_loocv_tr, paste0("model_predictions_loocv_tr.csv"))
  449. RMSE_subset_te <- data.frame(
  450. RMSE = RMSE(model_predictions_loocv_te, testing_data$Age),
  451. Rsquare = R2(model_predictions_loocv_te, testing_data$Age),
  452. MAE = mae(testing_data$Age, model_predictions_loocv_te),
  453. PEARSON = cor(testing_data$Age, model_predictions_loocv_te, method = 'pearson')
  454. )
  455. write.csv(RMSE_subset_te, paste0("RMSE_loocv_final_te.csv"))
  457. RMSE_subset_tr <- data.frame(
  458. RMSE = RMSE(model_predictions_loocv_tr, training_data$Age),
  459. Rsquare = R2(model_predictions_loocv_tr, training_data$Age),
  460. MAE = mae(training_data$Age, model_predictions_loocv_tr),
  461. PEARSON = cor(training_data$Age, model_predictions_loocv_tr, method = 'pearson')
  462. )
  463. write.csv(RMSE_subset_tr, paste0("RMSE_loocv_final_tr.csv"))
  466. ## MAE for both datasets
  468. setwd("C:/Users/jbelik/Documents/Assistanat/These/1-epigenetic clock/Article")
  470. abs <- read.table(file = "absolute error.csv", header = T, sep = ";", dec = ",")
  472. library(ggplot2)
  474. abs$Data <- factor(abs$Data, levels = c("Training", "Testing"))
  476. ggplot(abs, aes(x = Data, y = Absolute.Error, color = Data)) +
  477. geom_boxplot() +
  478. geom_point()+
  479. ylim(0, 150) +
  480. scale_color_manual(values = c("Testing" = "red", "Training"="grey")) +
  481. theme_minimal() +
  482. xlab("Samples") +
  483. ylab("Absolute Error (days)") +
  484. ggtitle("Absolute error in the training and testing datasets")+
  485. theme(legend.position="none")
  487. training <- abs[abs$Data=="Training",2]
  488. testing <- abs[abs$Data=="Testing",2]
  490. t.test(training, testing, alternative = "two.sided", var.equal = FALSE)

clock-young-boucle-file.R at commit 1486922, no license · at the source

Overview

Authors: Justine Bélik1, Frédéric Silvestre1
ORCID iDs: Justine Bélik
  1. Laboratory of Evolutionary and Adaptive Physiology, Institute of Life, Earth and Environment, University of Namur, Namur, Belgium
Institutions: University of Namur (Belgium); Institute of Life, Earth and Environment
Journal: Ecology and evolution, volume 16, issue 6, article e73881
Dates: received 23 February 2026; accepted 11 June 2026; published online 21 June 2026
Type: Research article · Language: English
License: CC BY
Identifiers: DOI 10.1002/ece3.73881 · PMID 42333303 · PMCID PMC13283775 · OpenAlex W7165476904
Open access: gold, a free copy (OpenAlex)
Status: code verified
Categories: genetics / omics (modality), human (organism), cellular / molecular (subfield)
Methods: Statistics, Smoothing, state filtering, decompositions, Machine learning
Keywords: brain aging, DNA methylation, epigenetic clock, mangrove rivulus
Topic: Epigenetics and DNA Methylation (Molecular Biology, Biochemistry, Genetics and Molecular Biology), according to OpenAlex
Citations: cited by 1 paper (Europe PMC); 85 references in the paper

Abstract

DNA methylation changes predictably with age across taxa, but in most species, these patterns are confounded by genetic variation. As a result, age‐predictive methylation models have mostly been developed in genetically heterogeneous, cross‐fertilizing organisms, limiting inference about epigenetic aging per se. Disentangling epigenetic and genetic effects is therefore essential for understanding aging, adaptation, and evolution. Here, we exploit the mangrove rivulus (Kryptolebias marmoratus), one of only two known self‐fertilizing vertebrates (together with K. hermaphroditus), to examine epigenetic aging in a system of naturally occurring near‐isogenic individuals. Using reduced‐representation bisulfite sequencing of 89 brain samples spanning 60–1100 days of age, we identified 40 CpG sites whose methylation levels predict chronological age with high accuracy (R 2 > 0.96, Median Absolute Error of 28.7 days). These 40 age‐associated CpG sites were linked to nearby genes with known roles in cellular maintenance and neurodegeneration. These include genes implicated in aging and neurodegenerative processes across vertebrates, such as lamin‐A, the aryl hydrocarbon receptor, and genes associated with Alzheimer's disease in humans. By leveraging a self‐fertilizing vertebrate, this study demonstrates that DNA methylation undergoes consistent, age‐associated changes across the lifespan in the near absence of genetic variation. Our results establish self‐fertilizing vertebrates as powerful models for disentangling epigenetic aging from genetic effects and provide a foundation for comparative and evolutionary studies of aging.

Reproduced under the paper's license (CC BY), from the paper cited above.

Repositories

Its files are read in the Code ↔ Paper reader above, with 7 matches between paragraphs and lines of code.

jubelik/Epigenetic-clock

License: none: the authors keep all their rights
State: the link answers, verified on 27 September 2026
Evidence: files inventoried
Commit: 1486922e3b8f6b1e34ada29fc9db5d617c935f3a, 2 June 2026
Languages: R (2)
Size: 3 files, 2 scripts
Software Heritage: not archived
Found in: “Data Availability Statement”
Holds: README
Not found: license file, CITATION.cff, environment file, tests, continuous integration, documentation
Tools: caret (2 files), tidyverse (2 files), ggplot2 (1 file), ggpubr (1 file)
Availability: 1 check, the latest on 27 September 2026: the link answers
  • 27 September 2026: the link answers
3 files
At the source: github.com/jubelik/Epigenetic-clock

Zenodo 20491333

License: CC-BY-4.0
State: the link answers, verified on 27 September 2026
Evidence: files inventoried
Languages: R (2)
Size: 2 files, 2 scripts
Software Heritage: not checked
Found in: “Data Availability Statement”
Not found: README, license file, CITATION.cff, environment file, tests, continuous integration, documentation
Tools: caret (2 files), tidyverse (2 files), ggplot2 (1 file), ggpubr (1 file)
Availability: 1 check, the latest on 27 September 2026: the link answers (HTTP 200)
  • 27 September 2026: the link answers (HTTP 200)
2 files
At the source: doi.org/10.5281/zenodo.20491333

The paper's code and data availability statement is in the Data section.

Tracing map

Proposed by the machine: these links were found in the paper and verified at the source, without human review. The map will receive a Zenodo DOI once one of the paper's authors has validated it with their ORCID.

What the map holds:

  • 2 repositories of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 4 scripts, each with its path and the digest of its content;
  • 7 matches between paragraphs of the paper and lines of the code (method lexical-v1);
  • neither the text of the paper nor the code itself.

Its JSON (tracing-map.json) is deposited on Zenodo with its DOI once the map is validated.

Data

No dataset and no data link were found in the paper.

Data Availability Statement

The datasets generated and/or analyzed during the current study are available in the NCBI repository, under the ID BioProject ID PRJNA1331489. The code is available on the following GitHub page: https://github.com/jubelik/Epigenetic‐clock (https://github.com/jubelik/Epigenetic-clock) and Zenodo repository https://doi.org/10.5281/zenodo.20491333.

Reproduced under the paper's license (CC BY), from the paper cited above.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 27 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 2 authors, 4 keywords, 83 references.

Cite

This paper

Bélik, J., & Silvestre, F. (2026). Epigenetic Aging in Brain Tissue of the Self-Fertilizing Vertebrate, &lt;i&gt;Kryptolebias marmoratus&lt;/i&gt;. Ecology and evolution, 16(6), e73881. https://doi.org/10.1002/ece3.73881

BibTeX

@article{belik2026epigenetic,
author = {Bélik, Justine and Silvestre, Frédéric},
title = {{Epigenetic Aging in Brain Tissue of the Self-Fertilizing Vertebrate, \&lt;i\&gt;Kryptolebias marmoratus\&lt;/i\&gt;}},
journal = {Ecology and evolution},
year = {2026},
month = jun,
volume = {16},
number = {6},
pages = {e73881},
publisher = {Wiley},
issn = {2045-7758},
doi = {10.1002/ece3.73881},
url = {https://doi.org/10.1002/ece3.73881},
pmid = {42333303},
pmcid = {PMC13283775}
}

RIS

TY - JOUR
AU - Bélik, Justine
AU - Silvestre, Frédéric
TI - Epigenetic Aging in Brain Tissue of the Self-Fertilizing Vertebrate, &lt;i&gt;Kryptolebias marmoratus&lt;/i&gt;
T2 - Ecology and evolution
J2 - Ecol Evol
PY - 2026
DA - 2026/06/21
VL - 16
IS - 6
SP - e73881
SN - 2045-7758
PB - Wiley
DO - 10.1002/ece3.73881
UR - https://doi.org/10.1002/ece3.73881
LA - en
ER -

CSL-JSON

{
"id": "10.1002/ece3.73881",
"type": "article-journal",
"title": "Epigenetic Aging in Brain Tissue of the Self-Fertilizing Vertebrate, &lt;i&gt;Kryptolebias marmoratus&lt;/i&gt;",
"container-title": "Ecology and evolution",
"author": [
{
"family": "Bélik",
"given": "Justine"
},
{
"family": "Silvestre",
"given": "Frédéric"
}
],
"container-title-short": "Ecol Evol",
"volume": "16",
"issue": "6",
"page": "e73881",
"DOI": "10.1002/ece3.73881",
"PMID": "42333303",
"PMCID": "PMC13283775",
"ISSN": "2045-7758",
"publisher": "Wiley",
"URL": "https://doi.org/10.1002/ece3.73881",
"language": "en",
"issued": {
"date-parts": [
[
2026,
6,
21
]
]
}
}

The tracing map gets a citation of its own once an author has validated it and it has a DOI.

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:10.1016/j.isci.2026.116439 [code]
Decoding the role of transcriptomic clocks in the human prefrontal cortex.
Journal: iScience
In common: ggplot2, tidyverse, genetics / omics, 6 references
[2] doi:10.1007/s11357-026-02195-x [code]
The aging epigenome: integrative analyses reveal intersection with Alzheimer's disease.
Journal: GeroScience
In common: ggpubr, ggplot2, tidyverse, genetics / omics, 5 references
[3] doi:10.1371/journal.pbio.3003757 [code]
Cell type-agnostic transcriptomic signatures enable uniform comparisons of neural maturation.
Journal: PLoS biology
In common: caret, ggpubr, ggplot2, 1 other tool, genetics / omics, 3 references
[4] doi:10.1186/s13073-026-01698-8 [code]
From aging to Alzheimer's disease: concordant brain DNA methylation changes in late life.
Journal: Genome medicine
In common: ggpubr, ggplot2, tidyverse, genetics / omics, 4 references
[5] doi:10.1038/s41598-026-48613-0 [code]
An snRNA-seq aging clock for the fruit fly head sheds light on sex-biased aging.
Journal: Scientific reports
In common: genetics / omics, cellular / molecular, 6 references
[6] doi:10.1038/s44400-026-00074-y [code]
Methylomic signatures of tau and amyloid-beta in transgenic mouse models of Alzheimer's disease neuropathology.
Journal: NPJ dementia
In common: ggpubr, ggplot2, tidyverse, genetics / omics, cellular / molecular, 3 references
[7] doi:10.1038/s41514-026-00391-9 [code]
Region-specific transcriptional signatures of brain aging in the absence of neuropathology at the single-cell level.
Journal: npj aging
In common: ggpubr, ggplot2, tidyverse, genetics / omics, cellular / molecular, 3 references
[8] doi:10.1186/s12916-026-04869-x [code]
The tissue-specific effects of glucose-lowering drug targets on aging mediated through DNA methylation: a multi-omics genetic study.
Journal: BMC medicine
In common: ggplot2, tidyverse, genetics / omics, 3 references
[9] doi:10.1038/s41467-026-77170-3 [code]
DNA methylation profiling identifies long-range epigenetic silencing of clustered protocadherins as a key determinant of meningioma progression.
Journal: Nature communications
In common: caret, ggpubr, ggplot2, 1 other tool, genetics / omics, cellular / molecular
[10] doi:10.7717/peerj.21426 [code]
Integrated transcriptomic identification and validation reveal key autophagy-associated biomarkers in sleep deprivation.
Journal: PeerJ
In common: caret, ggpubr, ggplot2, 1 other tool, genetics / omics, cellular / molecular

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.