This script analyses changes in DNAm following the intervention, further adjusted for predicted endothelial nuclei proportions.
Load required packages
library(DNAmArray)
library(SummarizedExperiment)
library(tidyverse)
library(limma)
library(sva)
library(methylGSA)
library(IlluminaHumanMethylationEPICanno.ilm10b4.hg19)
library(minfi)
library(bacon)
library(lattice)
library(ggrepel)
library(irlba)
library(BiocParallel)
library(FDb.InfiniumMethylation.hg19)
library(ggfortify)
library(GenomicRanges)
library(ggrepel)
library(irlba)Load data
load("../GOTO_Data/GOTO_targets-filtered.Rdata")
load('../GOTO_Data/GOTO_methData-filtered.Rdata')Subset
methData <- methData[ , methData$tissue == 'muscle' & !is.na(methData$cc_musc_endo)]
targets <- targets %>% filter(tissue == "muscle" & !is.na(cc_musc_endo))
methData## class: SummarizedExperiment
## dim: 755777 148
## metadata(0):
## assays(1): beta
## rownames(755777): cg18478105 cg09835024 ... cg10633746 cg12623625
## rowData names(57): cpg chr ... MASK_extBase MASK_general
## colnames: NULL
## colData names(40): DNA_labnr IOP2_ID ... cc_musc_endo cc_musc_smoothComplete betas
complete_betas <- na.omit(assay(methData))
dim(complete_betas)## [1] 558443 148Calculate principal components
pc <- prcomp_irlba(
t(complete_betas),
n=5)Inspect
summary(pc)## Importance of components:
## PC1 PC2 PC3 PC4 PC5
## Standard deviation 14.4818 12.1996 8.78633 5.85444 5.43005
## Proportion of Variance 0.2146 0.1523 0.07899 0.03507 0.03017
## Cumulative Proportion 0.2146 0.3669 0.44584 0.48090 0.51107dim(pc$x)## [1] 148 5head(pc$x)## PC1 PC2 PC3 PC4 PC5
## [1,] 14.42926 6.866957 6.060301 0.7101864 3.830923
## [2,] 7.08136 10.118715 9.300618 10.0907113 2.492887
## [3,] 18.81138 4.559519 17.508802 5.0086925 -1.194359
## [4,] 4.50816 11.648587 17.079214 -1.9232631 3.230099
## [5,] 17.23604 3.745656 15.736912 0.8151135 -5.418460
## [6,] 32.76144 -2.790399 9.983066 8.5434184 -5.326081Merge
targets <- as.data.frame(colData(methData))
targets <- cbind(
targets,
pc$x)
colData(methData) <- DataFrame(targets)Formula
formula <- ~timepoint + age + sex + smoke + plate + array_row +
PC1 + PC2 + PC3 + PC4 + PC5 + cc_musc_endoDesign
design <- model.matrix(formula,
data=colData(methData))Correlation for random effect
dupcor <- duplicateCorrelation(assay(methData),
design,
block = colData(methData)$IOP2_ID)Fit models
fit <- lmFit(assay(methData), design,
block = colData(methData)$IOP2_ID,
correlation = dupcor$consensus.correlation)## Warning: Partial NA coefficients for 1222 probe(s)Save
coef <- fit$coefficients[, 2]
se <- fit$stdev.unscaled[, 2] * fit$sigma
tstat <- coef / se
pval <- 2 * pt(-abs(tstat), fit$df.residual)
n <- ncol(design) + fit$df.residualRun bacon to estimate bias and inflation of test statistics
bc <- bacon::bacon(teststatistics = tstat,
effectsizes = coef,
standarderrors = se,
verbose = TRUE)## Use multinomial weighted sampling...## threshold = -5.4009
## Starting values:
## p0 = 1.0000, p1 = 0.0000, p2 = 0.0000
## mu0 = 0.0178, mu1 = 5.9346, mu2 = -5.8990
## sigma0 = 1.0955, sigma1 = 1.0955, sigma2 = 1.0955bc## Bacon-object containing 1 set(s) of 755777 test-statistics.
## ...estimated bias: 0.02.
## ...estimated inflation: 1.1.
##
## Empirical null estimates are based on 5000 iterations with a burnin-period of 2000.bacon::inflation(bc)## sigma.0
## 1.087727bacon::bias(bc)## mu.0
## 0.01990327Save bacon adjusted p-values and t statistics
pval <- bacon::pval(bc)
tstat <- bacon::tstat(bc)Rerun bacon to check bias and inflation
bc <- bacon::bacon(teststatistics = tstat,
effectsizes = coef,
standarderrors = se,
verbose = TRUE)## Use multinomial weighted sampling...## threshold = -5.4009
## Starting values:
## p0 = 1.0000, p1 = 0.0000, p2 = 0.0000
## mu0 = -0.0020, mu1 = 5.4376, mu2 = -5.4415
## sigma0 = 1.0072, sigma1 = 1.0072, sigma2 = 1.0072bc## Bacon-object containing 1 set(s) of 755777 test-statistics.
## ...estimated bias: 0.00022.
## ...estimated inflation: 1.
##
## Empirical null estimates are based on 5000 iterations with a burnin-period of 2000.bacon::inflation(bc)## sigma.0
## 1.004958bacon::bias(bc)## mu.0
## 0.0002246736Adjust p-values
padj_fdr <- p.adjust(pval, method="fdr")Save results
limma_base <- data.frame(cpg = rownames(fit$coefficients),
beta = coef, SE = se,
p = pval, padj_fdr = padj_fdr,
t = tstat, N = n)Look at top CpGs
top_cpgs <- limma_base %>%
filter(padj_fdr <= 0.05) %>%
arrange(padj_fdr)
print(paste0("There are ", nrow(top_cpgs),
" significant CpGs in muscle after adjusting for predicted endothelial cells."))## [1] "There are 162 significant CpGs in muscle after adjusting for predicted endothelial cells."print(paste0(nrow(top_cpgs %>% filter(beta<0)),
" of these are hypomethylated, and ",
nrow(top_cpgs %>% filter(beta>0)),
" are hypermethylated."))## [1] "142 of these are hypomethylated, and 20 are hypermethylated."save(top_cpgs, file='../GOTO_Data/GOTO_results-top-muscle-adj.Rdata')
save(limma_base, file='../GOTO_Data/GOTO_results-full-muscle-adj.Rdata')sessionInfo()## R version 4.2.2 (2022-10-31)
## Platform: x86_64-pc-linux-gnu (64-bit)
## Running under: Rocky Linux 8.10 (Green Obsidian)
##
## Matrix products: default
## BLAS/LAPACK: /usr/lib64/libopenblas-r0.3.15.so
##
## locale:
## [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
## [3] LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8
## [5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8
## [7] LC_PAPER=en_US.UTF-8 LC_NAME=C
## [9] LC_ADDRESS=C LC_TELEPHONE=C
## [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
##
## attached base packages:
## [1] parallel stats4 stats graphics grDevices utils datasets
## [8] methods base
##
## other attached packages:
## [1] GEOquery_2.62.2
## [2] MuSiC_0.2.0
## [3] nnls_1.4
## [4] gplots_3.1.3
## [5] plotly_4.10.1
## [6] SeuratObject_4.1.3
## [7] Seurat_4.3.0
## [8] gridExtra_2.3
## [9] lattice_0.21-8
## [10] bacon_1.22.0
## [11] ellipse_0.4.5
## [12] methylGSA_1.12.0
## [13] sva_3.42.0
## [14] genefilter_1.76.0
## [15] mgcv_1.8-42
## [16] nlme_3.1-162
## [17] limma_3.54.2
## [18] lmerTest_3.1-3
## [19] lme4_1.1-30
## [20] IlluminaHumanMethylationEPICanno.ilm10b4.hg19_0.6.0
## [21] snpStats_1.44.0
## [22] survival_3.5-5
## [23] ggrepel_0.9.1
## [24] ggfortify_0.4.14
## [25] irlba_2.3.5.1
## [26] Matrix_1.5-4.1
## [27] omicsPrint_1.14.0
## [28] MASS_7.3-60
## [29] DNAmArray_2.0.0
## [30] pls_2.8-2
## [31] FDb.InfiniumMethylation.hg19_2.2.0
## [32] org.Hs.eg.db_3.14.0
## [33] TxDb.Hsapiens.UCSC.hg19.knownGene_3.2.2
## [34] GenomicFeatures_1.46.5
## [35] AnnotationDbi_1.56.2
## [36] IlluminaHumanMethylationEPICmanifest_0.3.0
## [37] minfi_1.40.0
## [38] bumphunter_1.36.0
## [39] locfit_1.5-9.8
## [40] iterators_1.0.14
## [41] foreach_1.5.2
## [42] Biostrings_2.62.0
## [43] XVector_0.34.0
## [44] SummarizedExperiment_1.24.0
## [45] Biobase_2.58.0
## [46] MatrixGenerics_1.10.0
## [47] matrixStats_1.0.0
## [48] GenomicRanges_1.46.1
## [49] GenomeInfoDb_1.34.9
## [50] IRanges_2.32.0
## [51] S4Vectors_0.36.2
## [52] BiocGenerics_0.44.0
## [53] BiocParallel_1.32.6
## [54] MethylAid_1.28.0
## [55] forcats_0.5.2
## [56] stringr_1.5.0
## [57] dplyr_1.1.3
## [58] purrr_0.3.4
## [59] readr_2.1.2
## [60] tidyr_1.2.1
## [61] tibble_3.2.1
## [62] ggplot2_3.4.3
## [63] tidyverse_1.3.2
## [64] rmarkdown_2.16
##
## loaded via a namespace (and not attached):
## [1] ica_1.0-3
## [2] Rsamtools_2.10.0
## [3] cinaR_0.2.3
## [4] lmtest_0.9-40
## [5] crayon_1.5.2
## [6] rhdf5filters_1.10.1
## [7] backports_1.4.1
## [8] reprex_2.0.2
## [9] GOSemSim_2.20.0
## [10] rlang_1.1.1
## [11] ROCR_1.0-11
## [12] readxl_1.4.1
## [13] SparseM_1.81
## [14] nloptr_2.0.3
## [15] filelock_1.0.2
## [16] rjson_0.2.21
## [17] bit64_4.0.5
## [18] glue_1.6.2
## [19] sctransform_0.3.5
## [20] rngtools_1.5.2
## [21] spatstat.sparse_3.0-1
## [22] mcmc_0.9-7
## [23] spatstat.geom_3.2-1
## [24] DOSE_3.20.1
## [25] haven_2.5.1
## [26] tidyselect_1.2.0
## [27] fitdistrplus_1.1-11
## [28] XML_3.99-0.14
## [29] zoo_1.8-12
## [30] GenomicAlignments_1.30.0
## [31] MatrixModels_0.5-1
## [32] xtable_1.8-4
## [33] magrittr_2.0.3
## [34] evaluate_0.21
## [35] cli_3.6.1
## [36] zlibbioc_1.44.0
## [37] miniUI_0.1.1.1
## [38] rstudioapi_0.14
## [39] doRNG_1.8.6
## [40] sp_1.6-1
## [41] MultiAssayExperiment_1.20.0
## [42] bslib_0.5.0
## [43] fastmatch_1.1-3
## [44] treeio_1.18.1
## [45] shiny_1.7.2
## [46] xfun_0.39
## [47] askpass_1.1
## [48] multtest_2.50.0
## [49] cluster_2.1.4
## [50] caTools_1.18.2
## [51] tidygraph_1.2.2
## [52] KEGGREST_1.34.0
## [53] quantreg_5.94
## [54] base64_2.0.1
## [55] ape_5.7-1
## [56] scrime_1.3.5
## [57] listenv_0.9.0
## [58] png_0.1-8
## [59] reshape_0.8.9
## [60] future_1.32.0
## [61] withr_2.5.0
## [62] bitops_1.0-7
## [63] ggforce_0.3.4
## [64] plyr_1.8.8
## [65] cellranger_1.1.0
## [66] coda_0.19-4
## [67] pillar_1.9.0
## [68] cachem_1.0.8
## [69] fs_1.6.2
## [70] clusterProfiler_4.2.2
## [71] DelayedMatrixStats_1.16.0
## [72] vctrs_0.6.3
## [73] ellipsis_0.3.2
## [74] generics_0.1.3
## [75] tools_4.2.2
## [76] munsell_0.5.0
## [77] tweenr_2.0.2
## [78] fgsea_1.20.0
## [79] DelayedArray_0.24.0
## [80] abind_1.4-5
## [81] fastmap_1.1.1
## [82] compiler_4.2.2
## [83] httpuv_1.6.11
## [84] rtracklayer_1.54.0
## [85] beanplot_1.3.1
## [86] MCMCpack_1.6-3
## [87] GenomeInfoDbData_1.2.9
## [88] edgeR_3.40.2
## [89] deldir_1.0-9
## [90] utf8_1.2.3
## [91] later_1.3.1
## [92] RobustRankAggreg_1.2.1
## [93] BiocFileCache_2.2.1
## [94] jsonlite_1.8.5
## [95] scales_1.2.1
## [96] pbapply_1.7-0
## [97] tidytree_0.4.0
## [98] sparseMatrixStats_1.10.0
## [99] lazyeval_0.2.2
## [100] promises_1.2.0.1
## [101] goftest_1.2-3
## [102] spatstat.utils_3.0-3
## [103] reticulate_1.30
## [104] htm2txt_2.2.2
## [105] nor1mix_1.3-0
## [106] cowplot_1.1.1
## [107] statmod_1.5.0
## [108] siggenes_1.68.0
## [109] Rtsne_0.16
## [110] downloader_0.4
## [111] uwot_0.1.14
## [112] igraph_1.4.3
## [113] HDF5Array_1.22.1
## [114] numDeriv_2016.8-1.1
## [115] yaml_2.3.7
## [116] htmltools_0.5.5
## [117] memoise_2.0.1
## [118] BiocIO_1.8.0
## [119] graphlayouts_0.8.1
## [120] quadprog_1.5-8
## [121] viridisLite_0.4.2
## [122] digest_0.6.31
## [123] assertthat_0.2.1
## [124] mime_0.12
## [125] rappdirs_0.3.3
## [126] RSQLite_2.2.17
## [127] yulab.utils_0.0.6
## [128] future.apply_1.11.0
## [129] data.table_1.14.8
## [130] blob_1.2.4
## [131] preprocessCore_1.60.2
## [132] splines_4.2.2
## [133] labeling_0.4.2
## [134] Rhdf5lib_1.20.0
## [135] illuminaio_0.40.0
## [136] googledrive_2.0.0
## [137] RaggedExperiment_1.18.0
## [138] RCurl_1.98-1.12
## [139] broom_1.0.1
## [140] hms_1.1.2
## [141] modelr_0.1.9
## [142] rhdf5_2.42.1
## [143] colorspace_2.1-0
## [144] aplot_0.1.7
## [145] sass_0.4.6
## [146] Rcpp_1.0.10
## [147] mclust_6.0.0
## [148] RANN_2.6.1
## [149] enrichplot_1.14.2
## [150] fansi_1.0.4
## [151] tzdb_0.4.0
## [152] parallelly_1.36.0
## [153] R6_2.5.1
## [154] grid_4.2.2
## [155] ggridges_0.5.4
## [156] lifecycle_1.0.3
## [157] curl_5.0.1
## [158] googlesheets4_1.0.1
## [159] minqa_1.2.5
## [160] leiden_0.4.3
## [161] jquerylib_0.1.4
## [162] DO.db_2.9
## [163] qvalue_2.26.0
## [164] RcppAnnoy_0.0.20
## [165] RColorBrewer_1.1-3
## [166] spatstat.explore_3.1-0
## [167] htmlwidgets_1.5.4
## [168] polyclip_1.10-4
## [169] biomaRt_2.50.3
## [170] missMethyl_1.28.0
## [171] shadowtext_0.1.2
## [172] timechange_0.2.0
## [173] gridGraphics_0.5-1
## [174] reactome.db_1.77.0
## [175] rvest_1.0.3
## [176] globals_0.16.2
## [177] openssl_2.0.6
## [178] spatstat.random_3.1-5
## [179] patchwork_1.1.2
## [180] progressr_0.13.0
## [181] codetools_0.2-19
## [182] IlluminaHumanMethylation450kanno.ilmn12.hg19_0.6.0
## [183] lubridate_1.9.2
## [184] GO.db_3.14.0
## [185] gtools_3.9.4
## [186] prettyunits_1.1.1
## [187] dbplyr_2.2.1
## [188] gridBase_0.4-7
## [189] gtable_0.3.3
## [190] DBI_1.1.3
## [191] tensor_1.5
## [192] ggfun_0.0.7
## [193] httr_1.4.6
## [194] highr_0.10
## [195] KernSmooth_2.23-21
## [196] stringi_1.7.12
## [197] vroom_1.5.7
## [198] progress_1.2.2
## [199] reshape2_1.4.4
## [200] farver_2.1.1
## [201] annotate_1.72.0
## [202] viridis_0.6.2
## [203] hexbin_1.28.3
## [204] ggtree_3.2.1
## [205] xml2_1.3.4
## [206] boot_1.3-28.1
## [207] restfulr_0.0.15
## [208] scattermore_0.8
## [209] ggplotify_0.1.0
## [210] bit_4.0.5
## [211] spatstat.data_3.0-1
## [212] scatterpie_0.1.8
## [213] ggraph_2.0.6
## [214] pkgconfig_2.0.3
## [215] gargle_1.5.0
## [216] knitr_1.43Clear
rm(list=ls())