Recent studies associated the differences of cell-type proportions
may be correlated to certain phenotypes, such as cancer. Therefore, the
demand for the development of computational methods to predict cell type
proportions increased. Hereby, we developed deconvR, a
collection of functions designed for analyzing deconvolution of the bulk
sample(s) using an atlas of reference omic signature profiles and a
user-selected model. We wanted to give users an option to extend their
reference atlas. Users can create new reference atlases using
findSignatures or extend their atlas by adding more cell
types. Additionally, we included BSMeth2Probe to make
mapping whole-genome bisulfite sequencing data to their probe IDs
easier. So users can map WGBS methylation data (as in
methylKit or GRanges object format) to
probe IDs, and the results of this mapping can be used as the bulk
samples in the deconvolution. We also included a comprehensive DNA
methylation atlas of 25 different cell types to use in the main function
deconvolute. deconvolute allows the user to
specify their desired deconvolution model (non-negative least squares
regression, support vector regression, quadratic programming, or robust
linear regression), and returns a dataframe which contains predicted
cell-type proportions of bulk methylation profiles, as well as partial
R-squared values for each sample.
As an another option, users can generate a simulated table of a
desired number of samples, with either user-specified or random origin
proportions using simulateCellMix.
simulateCellMix returns a second data frame called
proportions, which contains the actual cell-type
proportions of the simulated sample. It can be used for testing the
accuracy of the deconvolution by comparing these actual proportions to
the proportions predicted by deconvolute.
deconvolute returns partial R-squares, to check if
deconvolution brings advantages on top of the basic bimodal profiles.
The reference matrix usually follows a bimodal distribution in the case
of methylation, and taking the average of the rows of methylation matrix
might give a pretty similar profile to the bulk methylation profile you
are trying to deconvolute. If the deconvolution is advantageous, partial
R-squared is expect to be high.
The deconvR package can be installed from Bioconductor with:
The comprehensive human methylome reference atlas created by Moss et
al. 1 can
be used as the reference atlas parameter for several functions in this
package. This atlas was modified to remove duplicate CpG loci before
being included in the package as the dataframe. The dataframe is
composed of 25 human cell types and roughly 6000 CpG loci, identified by
their Illumina Probe ID. For each cell type and CpG locus, a methylation
value between 0 and 1 is provided. This value represents the fraction of
methylated bases of the CpG locus. The atlas therefore provides a unique
methylation pattern for each cell type and can be directly used as
reference in deconvolute and
simulateCellMix, and atlas in
findSignatures. Below is an example dataframe to illustrate
the atlas format.
## IDs Monocytes_EPIC B.cells_EPIC CD4T.cells_EPIC NK.cells_EPIC
## 1 cg08169020 0.8866 0.2615 0.0149 0.0777
## 2 cg25913761 0.8363 0.2210 0.2816 0.4705
## 3 cg26955540 0.7658 0.0222 0.1492 0.4005
## 4 cg25170017 0.8861 0.5116 0.1021 0.4363
## 5 cg12827637 0.5212 0.3614 0.0227 0.2120
## 6 cg19442545 0.2013 0.1137 0.0608 0.0410The GRanges object
IlluminaMethEpicB5ProbeIDs contains the Illumina probe IDs
of 400000 genomic loci (identified using the “seqnames”, “ranges”, and
“strand” values). This object is based off of the Infinium
MethylationEPIC v1.0 B5 Manifest data. Unnecessary columns were removed
and rows were truncated to reduce file size before converting the file
to a GRanges object. It can be used directly as
probe_id_locations in BSmeth2Probe.
## GRanges object with 6 ranges and 1 metadata column:
## seqnames ranges strand | ID
## <Rle> <IRanges> <Rle> | <character>
## [1] chr19 5236005-5236006 + | cg07881041
## [2] chr20 63216298-63216299 + | cg18478105
## [3] chr1 6781065-6781066 + | cg23229610
## [4] chr2 197438742-197438743 - | cg03513874
## [5] chrX 24054523-24054524 + | cg09835024
## [6] chr14 93114794-93114795 - | cg05451842
## -------
## seqinfo: 25 sequences from an unspecified genome; no seqlengthsAs mentioned in the introduction section, users can extend their
reference atlas to incorporate new data. Or may combine different
reference atlases to construct a more comprehensive one. This can be
done using the findSignatures function. In this example,
since we don’t have any additional reference atlas, we will add
simulated data as a new cell type to reference atlas for example
purposes. First, ensure that atlas (the signature matrix to
be extended) and samples (the new data to be added to the
signature matrix) are compliant with the function requirements. Below
illustrates the samples format.
## IDs Sample 1 Sample 2 Sample 3
## 1 cg08169020 0.22046597 0.28610457 0.06461126
## 2 cg25913761 0.33561110 0.42366696 0.21627701
## 3 cg26955540 0.22462290 0.34036200 0.08723216
## 4 cg25170017 0.25340882 0.31787792 0.11591903
## 5 cg12827637 0.17330681 0.23842003 0.06189597
## 6 cg19442545 0.07453234 0.07144488 0.04316458sampleMeta should include all sample names in
samples, and specify the origins they should be mapped to
when added to atlas.
sampleMeta <- data.table("Experiment_accession" = colnames(samples)[-1],
"Biosample_term_name" = "new cell type")
head(sampleMeta)## Experiment_accession Biosample_term_name
## <char> <char>
## 1: Sample 1 new cell type
## 2: Sample 2 new cell type
## 3: Sample 3 new cell typeUse findSignatures to extend the matrix.
extended_matrix <- findSignatures(samples = samples,
sampleMeta = sampleMeta,
atlas = HumanCellTypeMethAtlas,
IDs = "IDs")## CELL TYPES IN EXTENDED ATLAS:
## new cell type
## Monocytes_EPIC
## B.cells_EPIC
## CD4T.cells_EPIC
## NK.cells_EPIC
## CD8T.cells_EPIC
## Neutrophils_EPIC
## Erythrocyte_progenitors
## Adipocytes
## Cortical_neurons
## Hepatocytes
## Lung_cells
## Pancreatic_beta_cells
## Pancreatic_acinar_cells
## Pancreatic_duct_cells
## Vascular_endothelial_cells
## Colon_epithelial_cells
## Left_atrium
## Bladder
## Breast
## Head_and_neck_larynx
## Kidney
## Prostate
## Thyroid
## Upper_GI
## Uterus_cervix## IDs new_cell_type Monocytes_EPIC B.cells_EPIC CD4T.cells_EPIC
## 1 cg08169020 0.19039393 0.8866 0.2615 0.0149
## 2 cg25913761 0.32518502 0.8363 0.2210 0.2816
## 3 cg26955540 0.21740568 0.7658 0.0222 0.1492
## 4 cg25170017 0.22906859 0.8861 0.5116 0.1021
## 5 cg12827637 0.15787427 0.5212 0.3614 0.0227
## 6 cg19442545 0.06304727 0.2013 0.1137 0.0608
## NK.cells_EPIC CD8T.cells_EPIC Neutrophils_EPIC Erythrocyte_progenitors
## 1 0.0777 0.0164 0.8680 0.9509
## 2 0.4705 0.3961 0.8293 0.2385
## 3 0.4005 0.3474 0.7915 0.1374
## 4 0.4363 0.0875 0.7042 0.9447
## 5 0.2120 0.0225 0.5368 0.4667
## 6 0.0410 0.0668 0.1952 0.1601
## Adipocytes Cortical_neurons Hepatocytes Lung_cells Pancreatic_beta_cells
## 1 0.0336 0.0168 0.0340 0.0416 0.038875
## 2 0.3578 0.3104 0.2389 0.2250 0.132000
## 3 0.1965 0.0978 0.0338 0.0768 0.041725
## 4 0.0842 0.2832 0.2259 0.0544 0.111750
## 5 0.0287 0.1368 0.0307 0.1607 0.065975
## 6 0.0364 0.0222 0.1574 0.0122 0.003825
## Pancreatic_acinar_cells Pancreatic_duct_cells Vascular_endothelial_cells
## 1 0.0209 0.0130 0.0323
## 2 0.2249 0.1996 0.3654
## 3 0.0314 0.0139 0.2382
## 4 0.0309 0.0217 0.0972
## 5 0.0370 0.0230 0.0798
## 6 0.0378 0.0347 0.0470
## Colon_epithelial_cells Left_atrium Bladder Breast Head_and_neck_larynx Kidney
## 1 0.0163 0.0386 0.0462 0.0264 0.0470 0.0269
## 2 0.2037 0.2446 0.2054 0.1922 0.2045 0.1596
## 3 0.0193 0.1134 0.1269 0.1651 0.1523 0.1034
## 4 0.0187 0.0674 0.0769 0.0691 0.0704 0.0604
## 5 0.0193 0.0432 0.0459 0.0228 0.0687 0.0234
## 6 0.0193 0.0287 0.0246 0.0081 0.0098 0.0309
## Prostate Thyroid Upper_GI Uterus_cervix
## 1 0.0353 0.0553 0.0701 0.0344
## 2 0.1557 0.1848 0.1680 0.2026
## 3 0.0686 0.0943 0.1298 0.1075
## 4 0.0369 0.0412 0.0924 0.0697
## 5 0.0508 0.0726 0.0759 0.0196
## 6 0.0055 0.0188 0.0090 0.0166WGBS methylation data first needs to be mapped to probes using
BSmeth2Probe before being deconvoluted. The methylation
data WGBS_data in BSmeth2Probe may be either a
GRanges object or a methylKit
object.
Format of WGBS_data as GRanges
object:
## GRanges object with 6 ranges and 1 metadata column:
## seqnames ranges strand | sample1
## <Rle> <IRanges> <Rle> | <numeric>
## [1] chr1 47176 + | 0.833333
## [2] chr1 47177 - | 0.818182
## [3] chr1 47205 + | 0.681818
## [4] chr1 47206 - | 0.583333
## [5] chr1 47362 + | 0.416667
## [6] chr1 49271 + | 0.733333
## -------
## seqinfo: 1 sequence from an unspecified genome; no seqlengthsor as methylKit object:
head(methylKit::methRead(system.file("extdata", "test1.myCpG.txt",
package = "methylKit"),
sample.id="test", assembly="hg18",
treatment=1, context="CpG", mincov = 0))## chr start end strand coverage numCs numTs
## 1 chr21 9764513 9764513 - 12 0 12
## 2 chr21 9764539 9764539 - 12 3 9
## 3 chr21 9820622 9820622 + 13 0 13
## 4 chr21 9837545 9837545 + 11 0 11
## 5 chr21 9849022 9849022 + 124 90 34
## 6 chr21 9853296 9853296 + 17 10 7probe_id_locations contains information needed to map
cellular loci to probe IDs
## GRanges object with 6 ranges and 1 metadata column:
## seqnames ranges strand | ID
## <Rle> <IRanges> <Rle> | <character>
## [1] chr19 5236005-5236006 + | cg07881041
## [2] chr20 63216298-63216299 + | cg18478105
## [3] chr1 6781065-6781066 + | cg23229610
## [4] chr2 197438742-197438743 - | cg03513874
## [5] chrX 24054523-24054524 + | cg09835024
## [6] chr14 93114794-93114795 - | cg05451842
## -------
## seqinfo: 25 sequences from an unspecified genome; no seqlengthsUse BSmeth2Probe to map WGBS data to probe IDs.
mapped_WGBS_data <- BSmeth2Probe(probe_id_locations = IlluminaMethEpicB5ProbeIDs,
WGBS_data = WGBS_GRanges,
multipleMapping = TRUE,
cutoff = 10)
head(mapped_WGBS_data)## IDs sample1
## 1 cg00305774 1.0000000
## 2 cg00546078 0.8181818
## 3 cg00546307 0.5454545
## 4 cg00546971 0.5000000
## 5 cg00774867 0.8461538
## 6 cg00830435 0.9166667This mapped data can now be used in deconvolute. Here we
will deconvolute it using the previously extended signature matrix as
the reference atlas.
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.9963 0.9963 0.9963 0.9963 0.9963 0.9963## Monocytes_EPIC B.cells_EPIC CD4T.cells_EPIC NK.cells_EPIC
## sample1 0 0 0 0
## CD8T.cells_EPIC Neutrophils_EPIC Erythrocyte_progenitors Adipocytes
## sample1 0 0 0 0.5011789
## Cortical_neurons Hepatocytes Lung_cells Pancreatic_beta_cells
## sample1 0.2917357 0 0 0
## Pancreatic_acinar_cells Pancreatic_duct_cells
## sample1 0 0
## Vascular_endothelial_cells Colon_epithelial_cells Left_atrium Bladder
## sample1 0 0.001012524 0 0
## Breast Head_and_neck_larynx Kidney Prostate Thyroid Upper_GI
## sample1 0 0.2060729 0 0 0 0
## Uterus_cervix
## sample1 0Alternatively, users can set tissueSpecCpGs as TRUE to construct tissue based methylation signature matrix by using the reference atlas. Here, we used simulated samples to construct tissue specific signature matrix since we don’t have tissue specific samples.
data("HumanCellTypeMethAtlas")
exampleSamples <- simulateCellMix(1,reference = HumanCellTypeMethAtlas)$simulated
exampleMeta <- data.table("Experiment_accession" = "example_sample",
"Biosample_term_name" = "example_cell_type")
colnames(exampleSamples) <- c("CpGs", "example_sample")
colnames(HumanCellTypeMethAtlas)[1] <- c("CpGs")
signatures <- findSignatures(
samples = exampleSamples,
sampleMeta = exampleMeta,
atlas = HumanCellTypeMethAtlas,
IDs = "CpGs", K = 100, tissueSpecCpGs = TRUE)## CELL TYPES IN EXTENDED ATLAS:
## example_cell_type
## Monocytes_EPIC
## B.cells_EPIC
## CD4T.cells_EPIC
## NK.cells_EPIC
## CD8T.cells_EPIC
## Neutrophils_EPIC
## Erythrocyte_progenitors
## Adipocytes
## Cortical_neurons
## Hepatocytes
## Lung_cells
## Pancreatic_beta_cells
## Pancreatic_acinar_cells
## Pancreatic_duct_cells
## Vascular_endothelial_cells
## Colon_epithelial_cells
## Left_atrium
## Bladder
## Breast
## Head_and_neck_larynx
## Kidney
## Prostate
## Thyroid
## Upper_GI
## Uterus_cervix## $hyper
## cg08169020 cg25170017 cg26955540 cg12827637 cg15838173 cg04858631 cg19442545
## 0.2350464 0.1676764 0.1670077 0.1617190 0.1474864 0.1454226 0.1436747
## cg10560079 cg00982136 cg22017733 cg26460530 cg24082121 cg13677741 cg09642825
## 0.1355211 0.1316909 0.1298737 0.1286474 0.1255213 0.1251204 0.1232391
## cg01424562 cg15376996 cg14789659 cg19300307 cg18856478 cg14785464 cg06340704
## 0.1215070 0.1204296 0.1177056 0.1166839 0.1165306 0.1165004 0.1163029
## cg12474798 cg11895835 cg26033520 cg27189395 cg05056497 cg05258935 cg06167719
## 0.1155638 0.1154137 0.1152528 0.1146239 0.1139271 0.1135101 0.1132498
## cg05507654 cg20684110 cg00936790 cg08474651 cg08857351 cg24275356 cg14855367
## 0.1130509 0.1124435 0.1121967 0.1121693 0.1118520 0.1115099 0.1112162
## cg25913761 cg27395200 cg23220823 cg22879098 cg10509607 cg08425810 cg08063160
## 0.1110516 0.1109643 0.1107833 0.1103709 0.1103256 0.1102918 0.1102562
## cg03254916 cg22897141 cg27388962 cg03711944 cg23247274 cg02962602 cg10361922
## 0.1098989 0.1098782 0.1092117 0.1090746 0.1089722 0.1088976 0.1082737
## cg08257293 cg09267773 cg08832851 cg26757820 cg04913246 cg26953232 cg26047334
## 0.1081797 0.1079160 0.1078965 0.1078616 0.1077810 0.1075636 0.1074475
## cg16416715 cg14845962 cg14057303 cg25158622 cg16127573 cg08367326 cg27312961
## 0.1074304 0.1069937 0.1065205 0.1062310 0.1061525 0.1058592 0.1058540
## cg15014975 cg11975790 cg16278496 cg08846870 cg18675610 cg19331221 cg05043794
## 0.1054016 0.1051418 0.1045768 0.1043609 0.1041936 0.1041708 0.1041708
## cg09577804 cg07865091 cg16402452 cg04055490 cg24591770 cg02297709 cg03063309
## 0.1038559 0.1037965 0.1036543 0.1036539 0.1035739 0.1035109 0.1034758
## cg11532431 cg22138735 cg21181453 cg10661769 cg06889535 cg04462378 cg06398251
## 0.1033835 0.1033617 0.1031846 0.1030278 0.1027390 0.1026528 0.1023721
## cg00769843 cg15185986 cg01377358 cg16017089 cg17847861 cg16306706 cg12461469
## 0.1022756 0.1017510 0.1014651 0.1013992 0.1011996 0.1011574 0.1011538
## cg15575375 cg11025609 cg15059065 cg07218880 cg08659179 cg07438103 cg16829306
## 0.1011498 0.1010282 0.1009826 0.1006366 0.1003984 0.1002468 0.1002454
## cg16988611 cg13638257
## 0.1002116 0.1000065
##
## $hypo
## cg03663120 cg20942286 cg03963853 cg03126713 cg24788483 cg00828556 cg11186858
## 0.3689695 0.2939045 0.2710877 0.2684766 0.2663304 0.2659268 0.2618581
## cg13931640 cg05612654 cg22528270 cg15871206 cg15310871 cg13500029 cg10480329
## 0.2614733 0.2601873 0.2589226 0.2584671 0.2561083 0.2550339 0.2466639
## cg11231069 cg12655112 cg03549146 cg12458039 cg05923857 cg07737292 cg20610950
## 0.2456836 0.2364590 0.2362169 0.2357754 0.2356890 0.2310415 0.2309331
## cg10456459 cg03313271 cg06988336 cg16636767 cg06517984 cg06297318 cg27334271
## 0.2273215 0.2215700 0.2184522 0.2122077 0.2117664 0.2105619 0.2090778
## cg14976569 cg04851465 cg11597902 cg23952578 cg15633603 cg11327657 cg22259797
## 0.2083122 0.2082104 0.2075830 0.2072294 0.2071989 0.2066873 0.2065451
## cg08425796 cg06373940 cg13403369 cg09322573 cg18990407 cg04354689 cg20107506
## 0.2024776 0.2022143 0.2018694 0.2013115 0.2007050 0.2003369 0.2000488
## cg22185977 cg07033722 cg06125903 cg25517015 cg10718056 cg01879591 cg26538782
## 0.1995268 0.1989977 0.1989565 0.1987006 0.1986858 0.1980379 0.1975884
## cg12866960 cg27366072 cg03310874 cg02796279 cg19502671 cg08538581 cg06978145
## 0.1975592 0.1959536 0.1946131 0.1946042 0.1931152 0.1928923 0.1921243
## cg13093111 cg26923863 cg10967114 cg23403750 cg06721411 cg26783127 cg01024962
## 0.1915503 0.1912083 0.1910990 0.1909644 0.1907709 0.1905686 0.1901840
## cg20429104 cg20966357 cg26305504 cg11153071 cg06585734 cg19380303 cg23755933
## 0.1900984 0.1896101 0.1892500 0.1891926 0.1890469 0.1890051 0.1888823
## cg04664897 cg12555086 cg24250902 cg07918933 cg04217515 cg25596405 cg27340480
## 0.1887980 0.1882581 0.1879459 0.1876854 0.1876844 0.1871776 0.1870987
## cg18835472 cg23334433 cg05445326 cg04836151 cg26298914 cg02244028 cg26889953
## 0.1867587 0.1863246 0.1861762 0.1858981 0.1858499 0.1849684 0.1849121
## cg08428868 cg17086773 cg14189391 cg15288326 cg00009088 cg11802666 cg22875823
## 0.1848778 0.1844473 0.1844372 0.1844354 0.1842986 0.1840982 0.1840229
## cg13980609 cg04586126 cg08400494 cg07417146 cg17117981 cg01622606 cg22666015
## 0.1840206 0.1836747 0.1835193 0.1834423 0.1824030 0.1817908 0.1815471
## cg00235484 cg22662844
## 0.1813682 0.1810431Alternatively, users can set tissueSpecDMPs as TRUE to obtain tissue based DMPs by using the reference atlas. Here, we used simulated samples since we don’t have tissue specific samples. Note that both tissueSpecCpGs and tissueSpecDMPs can’t be TRUE at the same time.
data("HumanCellTypeMethAtlas")
exampleSamples <- simulateCellMix(1,reference = HumanCellTypeMethAtlas)$simulated
exampleMeta <- data.table("Experiment_accession" = "example_sample",
"Biosample_term_name" = "example_cell_type")
colnames(exampleSamples) <- c("CpGs", "example_sample")
colnames(HumanCellTypeMethAtlas)[1] <- c("CpGs")
signatures <- findSignatures(
samples = exampleSamples,
sampleMeta = exampleMeta,
atlas = HumanCellTypeMethAtlas,
IDs = "CpGs", tissueSpecDMPs = TRUE)## CELL TYPES IN EXTENDED ATLAS:
## example_cell_type
## Monocytes_EPIC
## B.cells_EPIC
## CD4T.cells_EPIC
## NK.cells_EPIC
## CD8T.cells_EPIC
## Neutrophils_EPIC
## Erythrocyte_progenitors
## Adipocytes
## Cortical_neurons
## Hepatocytes
## Lung_cells
## Pancreatic_beta_cells
## Pancreatic_acinar_cells
## Pancreatic_duct_cells
## Vascular_endothelial_cells
## Colon_epithelial_cells
## Left_atrium
## Bladder
## Breast
## Head_and_neck_larynx
## Kidney
## Prostate
## Thyroid
## Upper_GI
## Uterus_cervix## intercept f pval qval
## cg10480329 -3.287511 215.8571 1.704004e-13 1.027173e-09
## cg06297318 -3.496336 183.2111 9.941648e-13 2.996413e-09
## cg18835472 -3.350244 155.6614 5.558858e-12 1.116960e-08
## cg05923857 -3.515739 147.9330 9.451719e-12 1.406518e-08
## cg27366072 -2.858145 144.9613 1.166654e-11 1.406518e-08
## cg15633603 -1.496293 141.5312 1.494547e-11 1.501522e-08It is possible to use RNA-seq data for deconvolution via
deconvR package. Beware that you have to set
IDs column that contains Gene names to run
deconvR functions. Therefore you can simulate bulk
RNA-seq data via simulateCellMix and, extend RNA-seq
reference atlas via findSignatures.
## R version 4.4.1 (2024-06-14)
## Platform: x86_64-pc-linux-gnu
## Running under: Ubuntu 24.04.1 LTS
##
## Matrix products: default
## BLAS: /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3
## LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.26.so; LAPACK version 3.12.0
##
## locale:
## [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
## [3] LC_TIME=en_US.UTF-8 LC_COLLATE=C
## [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
##
## time zone: Etc/UTC
## tzcode source: system (glibc)
##
## attached base packages:
## [1] parallel stats graphics grDevices utils datasets methods
## [8] base
##
## other attached packages:
## [1] dplyr_1.1.4 doParallel_1.0.17 iterators_1.0.14 foreach_1.5.2
## [5] deconvR_1.13.0 data.table_1.16.2 knitr_1.48 BiocStyle_2.35.0
##
## loaded via a namespace (and not attached):
## [1] splines_4.4.1 BiocIO_1.17.0
## [3] bitops_1.0-9 tibble_3.2.1
## [5] R.oo_1.26.0 preprocessCore_1.67.1
## [7] XML_3.99-0.17 lifecycle_1.0.4
## [9] lattice_0.22-6 MASS_7.3-61
## [11] base64_2.0.2 scrime_1.3.5
## [13] magrittr_2.0.3 minfi_1.51.0
## [15] limma_3.61.12 sass_0.4.9
## [17] rmarkdown_2.28 jquerylib_0.1.4
## [19] yaml_2.3.10 robslopes_1.1.3
## [21] doRNG_1.8.6 askpass_1.2.1
## [23] minqa_1.2.8 DBI_1.2.3
## [25] buildtools_1.0.0 RColorBrewer_1.1-3
## [27] abind_1.4-8 zlibbioc_1.51.2
## [29] quadprog_1.5-8 GenomicRanges_1.57.2
## [31] purrr_1.0.2 R.utils_2.12.3
## [33] BiocGenerics_0.53.0 RCurl_1.98-1.16
## [35] GenomeInfoDbData_1.2.13 IRanges_2.39.2
## [37] S4Vectors_0.43.2 maketools_1.3.1
## [39] rentrez_1.2.3 genefilter_1.87.0
## [41] annotate_1.85.0 DelayedMatrixStats_1.27.3
## [43] codetools_0.2-20 DelayedArray_0.31.14
## [45] xml2_1.3.6 tidyselect_1.2.1
## [47] UCSC.utils_1.1.0 lme4_1.1-35.5
## [49] beanplot_1.3.1 matrixStats_1.4.1
## [51] stats4_4.4.1 illuminaio_0.47.0
## [53] GenomicAlignments_1.41.0 jsonlite_1.8.9
## [55] multtest_2.61.0 e1071_1.7-16
## [57] survival_3.7-0 bbmle_1.0.25.1
## [59] tools_4.4.1 rsq_2.7
## [61] Rcpp_1.0.13 glue_1.8.0
## [63] SparseArray_1.5.45 xfun_0.48
## [65] qvalue_2.37.0 MatrixGenerics_1.17.1
## [67] GenomeInfoDb_1.41.2 HDF5Array_1.33.8
## [69] numDeriv_2016.8-1.1 BiocManager_1.30.25
## [71] fastmap_1.2.0 boot_1.3-31
## [73] rhdf5filters_1.17.0 fansi_1.0.6
## [75] openssl_2.2.2 digest_0.6.37
## [77] R6_2.5.1 colorspace_2.1-1
## [79] gtools_3.9.5 RSQLite_2.3.7
## [81] R.methodsS3_1.8.2 utf8_1.2.4
## [83] tidyr_1.3.1 generics_0.1.3
## [85] rtracklayer_1.65.0 class_7.3-22
## [87] httr_1.4.7 S4Arrays_1.5.11
## [89] pkgconfig_2.0.3 gtable_0.3.6
## [91] blob_1.2.4 siggenes_1.79.0
## [93] XVector_0.45.0 sys_3.4.3
## [95] htmltools_0.5.8.1 scales_1.3.0
## [97] Biobase_2.67.0 png_0.1-8
## [99] deming_1.4-1 tzdb_0.4.0
## [101] reshape2_1.4.4 rjson_0.2.23
## [103] nloptr_2.1.1 coda_0.19-4.1
## [105] nlme_3.1-166 curl_5.2.3
## [107] bdsmatrix_1.3-7 bumphunter_1.49.0
## [109] proxy_0.4-27 cachem_1.1.0
## [111] rhdf5_2.49.0 stringr_1.5.1
## [113] AnnotationDbi_1.69.0 restfulr_0.0.15
## [115] GEOquery_2.73.5 pillar_1.9.0
## [117] grid_4.4.1 reshape_0.8.9
## [119] vctrs_0.6.5 Deriv_4.1.6
## [121] xtable_1.8-4 evaluate_1.0.1
## [123] readr_2.1.5 GenomicFeatures_1.57.1
## [125] mvtnorm_1.3-1 cli_3.6.3
## [127] locfit_1.5-9.10 compiler_4.4.1
## [129] Rsamtools_2.21.2 rlang_1.1.4
## [131] crayon_1.5.3 rngtools_1.5.2
## [133] nor1mix_1.3-3 mclust_6.1.1
## [135] emdbook_1.3.13 plyr_1.8.9
## [137] stringi_1.8.4 mcr_1.3.3.1
## [139] BiocParallel_1.39.0 nnls_1.6
## [141] assertthat_0.2.1 munsell_0.5.1
## [143] Biostrings_2.75.0 Matrix_1.7-1
## [145] hms_1.1.3 sparseMatrixStats_1.17.2
## [147] bit64_4.5.2 ggplot2_3.5.1
## [149] Rhdf5lib_1.27.0 KEGGREST_1.45.1
## [151] methylKit_1.31.0 statmod_1.5.0
## [153] SummarizedExperiment_1.35.5 fastseg_1.51.0
## [155] memoise_2.0.1 bslib_0.8.0
## [157] bit_4.5.0Moss, J. et al. (2018). Comprehensive human cell-type methylation atlas reveals origins of circulating cell-free DNA in health and disease. Nature communications, 9(1), 1-12. https://doi.org/10.1038/s41467-018-07466-6↩︎