Last updated: 2020-01-26
Checks: 7 0
Knit directory: peco-paper/
This reproducible R Markdown analysis was created with workflowr (version 1.6.0). The Checks tab describes the reproducibility checks that were applied when the results were created. The Past versions tab lists the development history.
Great! Since the R Markdown file has been committed to the Git repository, you know the exact version of the code that produced these results.
Great job! The global environment was empty. Objects defined in the global environment can affect the analysis in your R Markdown file in unknown ways. For reproduciblity it’s best to always run the code in an empty environment.
The command set.seed(20190814) was run prior to running the code in the R Markdown file. Setting a seed ensures that any results that rely on randomness, e.g. subsampling or permutations, are reproducible.
Great job! Recording the operating system, R version, and package versions is critical for reproducibility.
Nice! There were no cached chunks for this analysis, so you can be confident that you successfully produced the results during this run.
Great job! Using relative paths to the files within your workflowr project makes it easier to run your code on other machines.
Great! You are using Git for version control. Tracking code development and connecting the code version to the results is critical for reproducibility. The version displayed above was the version of the Git repository at the time these results were generated.
Note that you need to be careful to ensure that all relevant files for the analysis have been committed to Git prior to generating the results (you can use wflow_publish or wflow_git_commit). workflowr only checks the R Markdown file, but you know if there are other scripts or data files that it depends on. Below is the status of the Git repository when the results were generated:
Ignored files:
Ignored: .Rhistory
Ignored: .Rproj.user/
Untracked files:
Untracked: analysis/eval_on_leng2015_data.Rmd
Untracked: analysis/npreg_trendfilter_quantile.Rmd
Untracked: data/HumanLengESC.rds
Untracked: data/data_training_test/
Untracked: data/eset-filtered.rds
Untracked: data/fit.quant.rds
Untracked: data/fit.trend.perm.lowmiss.rds
Untracked: data/fit_diff_cyclone.rds
Untracked: data/fit_diff_oscope.rds
Untracked: data/fit_diff_peco.rds
Untracked: data/fit_diff_recat.rds
Untracked: data/fit_diff_seurat.rds
Untracked: data/intensity.rds
Untracked: data/leng2015_data.rds
Untracked: data/leng_fucci_oscope_29genes.rda
Untracked: data/leng_fucci_recat.rda
Untracked: data/leng_geneinfo.txt
Untracked: data/log2cpm.quant.rds
Untracked: data/macosko-2015.rds
Untracked: data/nmeth.3549-S2.xlsx
Untracked: data/ourdata_cyclone_NA18511.rds
Untracked: data/ourdata_cyclone_NA18855.rds
Untracked: data/ourdata_cyclone_NA18870.rds
Untracked: data/ourdata_cyclone_NA19098.rds
Untracked: data/ourdata_cyclone_NA19101.rds
Untracked: data/ourdata_cyclone_NA19160.rds
Untracked: data/ourdata_oscope_366genes.rda
Untracked: data/ourdata_peco_NA18511_top005genes.rds
Untracked: data/ourdata_peco_NA18855_top005genes.rds
Untracked: data/ourdata_peco_NA18870_top005genes.rds
Untracked: data/ourdata_peco_NA19098_top005genes.rds
Untracked: data/ourdata_peco_NA19101_top005genes.rds
Untracked: data/ourdata_peco_NA19160_top005genes.rds
Untracked: data/ourdata_phase_cyclone.rds
Untracked: data/ourdata_phase_seurat.rds
Untracked: data/ourdata_recat.rda
Untracked: data/sce-filtered.rds
Unstaged changes:
Modified: analysis/index.Rmd
Modified: code/run_seurat.R
Note that any generated files, e.g. HTML, png, CSS, etc., are not included in this status report because it is ok for generated content to have uncommitted changes.
These are the previous versions of the R Markdown and HTML files. If you’ve configured a remote Git repository (see ?wflow_git_remote), click on the hyperlinks in the table below to view them.
| File | Version | Author | Date | Message |
|---|---|---|---|---|
| Rmd | c9afbb3 | jhsiao999 | 2020-01-26 | minor edits |
| html | 958a49a | jhsiao999 | 2020-01-25 | Build site. |
| Rmd | 2a13ee4 | jhsiao999 | 2020-01-25 | Comparisons of predictions on our data |
We compare the performance of peco with Seurat, Cyclone, Oscope and reCAT using our data in six-fold cross-validation. Here we provide the code that we used to predict continuous cel cycle phase in each method and to compute prediction error between FUCCI phase and predicted cell cycle phase.
Load packages
library(SingleCellExperiment)
library(peco)Prepare training/testing data
sce <- readRDS("data/sce-final.rds")
sce <- sce[grep("ENSG", rownames(sce)),]
fdata <- data.frame(colData(sce))
fdata <- data.frame(rowData(sce))
counts <- data.frame(assay(sce, "counts"))
sce_normed <- data_transform_quantile(sce)
log2cpm_quant <- assay(sce_normed, "cpm_quantNormed")
inds <- c("NA18511", "NA18855", "NA18870", "NA19098", "NA19101", "NA19160")
theta <- pdata$theta
# function to make training/testing data
makedata_supervised <- function(sce, log2cpm_quant,
theta) {
message("Create data/data_training_test folder \n")
if (!file.exists("data/data_training_test")) { dir.create("data/data_training_test") }
library(SingleCellExperiment)
library(peco)
pdata <- data.frame(colData(sce))
fdata <- data.frame(rowData(sce))
counts <- data.frame(assay(sce))
counts <- counts[grep("ENSG", rownames(counts)), ]
log2cpm <- t(log2(1+(10^6)*(t(counts)/pdata$molecules)))
for (ind in unique(pdata$chip_id)) {
ii_test <- c(1:nrow(pdata))[which(pdata$chip_id == ind)]
ii_train <- c(1:nrow(pdata))[which(pdata$chip_id != ind)]
pdata_test <- pdata[ii_test,]
pdata_train <- pdata[ii_train,]
log2cpm_quant_test <- log2cpm_quant[,ii_test]
log2cpm_quant_train <- log2cpm_quant[,ii_train]
theta <- pdata$theta
names(theta) <- rownames(pdata)
log2cpm_test <- log2cpm[,ii_test]
log2cpm_train <- log2cpm[,ii_train]
counts_test <- counts[,ii_test]
counts_train <- counts[,ii_train]
theta_test <- theta[ii_test]
theta_train <- theta[ii_train]
#sig.genes <- readRDS("output/npreg-trendfilter-quantile.Rmd/out.stats.ordered.sig.476.rds")
data_training <- list(theta_train=theta_train,
log2cpm_quant_train=log2cpm_quant_train,
log2cpm_train=log2cpm_train,
counts_train=counts_train,
pdata_train=pdata_train,
fdata=fdata)
data_test <- list(theta_test=theta_test,
log2cpm_quant_test=log2cpm_quant_test,
log2cpm_test=log2cpm_test,
counts_test = counts_test,
pdata_test=pdata_test,
fdata=fdata)
saveRDS(data_training,
file=file.path(paste0("data/data_training_test/ind_",ind,"_data_training.rds")))
saveRDS(data_test,
file=file.path(paste0("data/data_training_test/ind_",ind,"_data_test.rds")))
}
}
makedata_supervised(sce, log2cpm_quant, theta)Use peco predictor of 5 genes to predict cell cycle phase.
# --- prediction for samples from one individual cell ine
ngenes = 5
for (i in seq_along(inds)) {
ind = inds[i]
data_train <- readRDS(paste0("data/data_training_test/ind_", ind, "_data_training.rds"))
data_test <- readRDS(paste0("data/data_training_test/ind_",ind, "_data_test.rds"))
fits_all <- readRDS("data/fit.quant.rds")
genes_all <- names(fits_all)[order(sapply(fits_all,"[[",3), decreasing=T)]
which_genes <- genes_all[1:ngenes]
fit_train <- cycle_npreg_insample(
Y = with(data_train, log2cpm_quant_train[which(rownames(log2cpm_quant_train) %in% which_genes), ]),
theta = with(data_train, theta_train),
polyorder=2,
ncores=4,
method.trend="trendfilter")
fit_test <- cycle_npreg_outsample(
Y_test=with(data_test, log2cpm_quant_test[which(rownames(log2cpm_quant_test) %in% which_genes), ]),
sigma_est=with(fit_train, sigma_est),
funs_est=with(fit_train, funs_est),
method.grid = "uniform",
method.trend="trendfilter",
polyorder=2,
ncores=2)
out_peco <- list(fit_train=fit_train,
fit_test=fit_test)
saveRDS(out, file.path(paste0("data/ourdata_peco_",
ind, "_top", sprintf("%03d", ngenes),"genes.rds")))
}
# compute prediction error
inds <- c("NA19098","NA18511","NA18870","NA19101","NA18855","NA19160")
ngenes <- 5
diff_peco <- do.call(rbind, lapply(1:length(inds), function(i) {
ind <- inds[i]
fl_name <- list.files(file.path("data"),
pattern=paste0("ourdata_peco_", ind, "_top", sprintf("%03d", ngenes),"genes.rds"),
full.names = TRUE)
res_pred <- readRDS(fl_name)$fit_test
data_test <- readRDS(paste0("data/data_training_test/ind_",ind, "_data_test.rds"))
all.equal(rownames(data_test$pdata_test), names(res_pred$cell_times_est))
df_pred <- readRDS(fl_name)$fit_test
phase_ref = data_test$theta_test
phase_pred_rot = rotation(ref_var=data_test$theta_test,
shift_var=res_pred$cell_times_est)
return(data.frame(phase_ref = phase_ref,
phase_pred_rot = phase_pred_rot,
diff_time = circ_dist(phase_ref, phase_pred_rot),
ind = as.character(ind),
method = "peco"))
}))
diff_peco$ind = factor(diff_peco$ind,
levels = c("NA18511", "NA18855", "NA18870",
"NA19098", "NA19101", "NA19160"))
saveRDS(diff_peco, file=("data/fit_diff_peco.rds"))
diff_peco %>% group_by(ind) %>% summarise(mn=mean(diff_time)/2/pi,
sd=sd(diff_time/2/pi)/sqrt(sum(phase_pred_rot>0)))We computed Seurat cell cycle scores and transformed these two scores into continuous cell cycle phase. Specifically, we applied PCA to the two cell cycle scores and then assigend each cell an angle on a unit circle. We did this for samples of each individual cell lines separately.
# use Seurat to assign phase for each individual
# and then compute predictione error for samples in individual cell lines
diff_seurat <- do.call(rbind, lapply(1:length(inds), function(i) {
source("code/run_seurat.R")
seurat.genes <- readLines(con = "data/regev_lab_cell_cycle_genes.txt")
seurat.genes <- list(s.genes=seurat.genes[1:43],
g2m.genes=seurat.genes[44:97])
# print(i)
data <- readRDS(paste0("data/data_training_test/ind_",inds[i], "_data_test.rds"))
Y_fit <- data$log2cpm_test
rownames(Y_fit) <- fdata$name[match(rownames(Y_fit), rownames(fdata))]
fit <- run_seurat(Y_fit,
s.genes=seurat.genes$s.genes,
g2m.genes=seurat.genes$g2m.genes)
fit <- as.list(fit)
seurat.pca <- prcomp(cbind(fit$G2M, fit$S), scale=TRUE)
phase_pred <- intensity2circle(seurat.pca$x, plot.it = F, method = "trig")
names(phase_pred) <- colnames(Y_fit)
phase_ref <- pdata$theta[match(names(phase_pred), rownames(pdata))]
phase_pred_rot <- rotation(ref_var=phase_ref, shift_var=phase_pred)
return(data.frame(phase_ref = phase_ref,
phase_pred_rot = phase_pred_rot,
diff_time = circ_dist(phase_ref, phase_pred_rot),
ind = inds[i],
method = "Seurat"))
}) )
saveRDS(diff_seurat, file="data/fit_diff_seurat.rds")
diff_seurat %>% group_by(ind) %>% summarise(mn=mean(diff_time)/2/pi,
sd=sd(diff_time/2/pi)/sqrt(sum(phase_pred_rot>0)))We computed Cyclone cell cycle scores and transformed these two scores into continuous cell cycle phase. Specifically, we applied PCA to the two cell cycle scores and then assigend each cell an angle on a unit circle. We did this for samples of each individual cell lines separately.
library(scran)
hs.pairs <- readRDS(system.file("exdata", "human_cycle_markers.rds", package="scran"))
for (ind in unique(pData(eset)$chip_id)) {
data <- readRDS(file.path(paste0("data/data_training_test/ind_",
ind, "_data_test.rds")))
input <- data.frame(data$log2cpm_quant_test)
ii <- which(colSums(is.na(input))==0)
input <- input[,ii]
# --- begin cyclone
hs.pairs <- readRDS(system.file("exdata", "human_cycle_markers.rds", package="scran"))
# message("run cyclone", "\n")
out <- cyclone(input, pairs = hs.pairs,
gene.names=rownames(input),
iter=1000, min.iter=100, min.pairs=50,
BPPARAM=SerialParam(), verbose=T, subset.row=NULL)
out_cyclone <- data.frame(phase_cyclone = out$phases,
normalized.scores_G1 = out$normalized.scores$G1,
normalized.scores_S = out$normalized.scores$S,
normalized.scores_G2M = out$normalized.scores$G2M)
rownames(out_cyclone) <- colnames(input)
# Infer phase using Seurat
pc_cyclone <- prcomp(out$normalized.scores, scale. = T, center = T)
library(peco)
phase_peco <- intensity2circle(pc_cyclone$x[,1:2], plot.it = F, method = "trig")
out_cyclone$phase_pred <- rotation(data$theta_test,
phase_peco)
out_cyclone$phase_ref <- data$theta_test
out_cyclone$ind <- ind
saveRDS(out_cyclone, file = paste0("data/ourdata_cyclone_",
ind,".rds"))
}
# compute prediction error of cyclone results
inds <- c("NA19098","NA18511","NA18870","NA19101","NA18855","NA19160")
diff_cyclone <- do.call(rbind, lapply(1:length(inds), function(i) {
out_cyclone <- readRDS(paste0("data/ourdata_cyclone_",inds[i],".rds"))
phase_peco <- pdata$theta[match(rownames(out_cyclone), rownames(pdata))]
phase_cyclone_rot <- rotation(ref_var= phase_peco, shift_var = out_cyclone$phase_pred)
return(data.frame(phase_ref = phase_peco,
phase_pred_rot = phase_cyclone_rot,
diff_time = circ_dist(phase_peco, phase_cyclone_rot),
ind = inds[i],
method = "Cyclone"))
}) )
saveRDS(diff_cyclone, file="data/fit_diff_cyclone.rds")
diff_cyclone %>% group_by(ind) %>% summarise(mn=mean(diff_time)/2/pi,
sd=sd(diff_time/2/pi)/sqrt(sum(phase_pred_rot>0)))We applied Oscope to estimate cell cycle ordering across the 888 single-cell samples. The analysis used 366 genes that were selected using Oscope’s criterion of high variabilty genes. Specifically, we changed the default ’MeanCutlow` option of 100 to 10. We also found that the results were similar when adding more genes in the analysis.
library(Oscope)
# to run Oscope
# use gene symbols as labels
rownames(counts) <- fdata$name
Sizes <- MedianNorm(counts)
DataNorm <- GetNormalizedMat(counts, Sizes)
MV <- CalcMV(Data = as.matrix(counts), Sizes = Sizes, MeanCutLow = 10)
DataNorm_high_var <- DataNorm[MV$GeneToUse,]
DataInput <- NormForSine(DataNorm_high_var)
dim(DataInput)
# select samples from all individuals
SineRes <- OscopeSine(DataInput, parallel=T)
KMRes <- OscopeKM(SineRes, maxK = NULL)
# to flag clusters...
# no need to flag clusters
KMResUse <- KMRes
ENIRes <- OscopeENI(KMRes = KMResUse, Data = DataInput,
NCThre = 100, parallel=T)
save(DataInput,
KMResUse, ENIRes,
file = file.path("data/ourdata_oscope_366genes.rda")
# Oscope results
load("data/ourdata_oscope_366genes.rda")
samples_ordered <- colnames(DataInput)[ENIRes[["cluster2"]]]
phase_pred <- seq(0, 2*pi, length.out= length(ENIRes[["cluster2"]]))
names(phase_pred) <- samples_ordered
diff_oscope <- data.frame(phase_ref=pdata$theta[match(samples_ordered, colnames(eset))])
diff_oscope$phase_pred_rot <- rotation(ref_var= diff_oscope$phase_ref, shift_var = phase_pred)
diff_oscope$diff_time <- circ_dist(diff_oscope$phase_ref, diff_oscope$phase_pred_rot)
diff_oscope$ind <- pdata$chip_id[match(names(phase_pred), colnames(counts))]
diff_oscope$method <- "Oscope"
saveRDS(diff_oscope, file="data/fit_diff_oscope.rds")
diff_oscope %>% group_by(ind) %>% summarise(mn=mean(diff_time)/2/pi,
sd=sd(diff_time/2/pi)/sqrt(sum(phase_pred_rot>0)))We applied reCAT to estimate cell cycle ordering across the 888 single-cell samples. The analysis used all the 11,040 genes.
To run recAT, please clone the reCAT GitHUb repository and then cd to the directory.
git clone https://github.com/tinglab/reCAT
cd "reCAT/R"
# recat
input <- t(log2cpm_quant)
input <- input[sample(1:nrow(input)),]
source("get_test_exp.R")
test_exp <- get_test_exp(t(input))
source("get_ordIndex.R")
res_ord <- get_ordIndex(test_exp, 10)
ordIndex <- res_ord$ordIndex
save(test_exp, ordIndex,
file = file.path("data/ourdata_recat_rda"))
# compute predictor error of reCAT predicted time
sample_ordered <- rownames(test_exp)[ordIndex]
phase_pred <- seq(0, 2*pi, length.out= length(ordIndex))
names(phase_pred) <- sample_ordered
diff_recat <- data.frame(phase_ref=pdata$theta[match(samples_ordered, colnames(eset))])
diff_recat$ind <- pdata$chip_id[match(names(phase_pred), colnames(eset))]
diff_recat$phase_pred_rot <- rotation(ref_var= diff_recat$phase_ref, shift_var = phase_pred)
rownames(diff_recat) <- samples_ordered
diff_recat$diff_time <- circ_dist(diff_recat$phase_ref, diff_recat$phase_pred_rot)
diff_recat$method <- "reCAT"
diff_recat <- diff_recat[,c(1,3,4,2,5)]
saveRDS(diff_recat, file="data/fit_diff_recat.rds")
diff_recat %>% group_by(ind) %>% summarise(mn=mean(diff_time)/2/pi,
sd=sd(diff_time/2/pi)/sqrt(sum(phase_pred_rot>0)))sessionInfo()R version 3.5.1 (2018-07-02)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Scientific Linux 7.4 (Nitrogen)
Matrix products: default
BLAS/LAPACK: /software/openblas-0.2.19-el7-x86_64/lib/libopenblas_haswellp-r0.2.19.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] peco_0.99.6 SingleCellExperiment_1.4.1
[3] SummarizedExperiment_1.12.0 DelayedArray_0.8.0
[5] BiocParallel_1.16.0 matrixStats_0.55.0
[7] Biobase_2.42.0 GenomicRanges_1.34.0
[9] GenomeInfoDb_1.18.1 IRanges_2.16.0
[11] S4Vectors_0.20.1 BiocGenerics_0.28.0
loaded via a namespace (and not attached):
[1] viridis_0.5.1 genlasso_1.4
[3] viridisLite_0.3.0 foreach_1.4.4
[5] DelayedMatrixStats_1.4.0 assertthat_0.2.1
[7] vipor_0.4.5 GenomeInfoDbData_1.2.0
[9] yaml_2.2.0 pillar_1.3.1
[11] backports_1.1.2 lattice_0.20-38
[13] glue_1.3.0 digest_0.6.20
[15] promises_1.0.1 XVector_0.22.0
[17] colorspace_1.3-2 plyr_1.8.4
[19] htmltools_0.3.6 httpuv_1.4.5
[21] Matrix_1.2-17 pkgconfig_2.0.3
[23] zlibbioc_1.28.0 purrr_0.3.2
[25] mvtnorm_1.0-11 scales_1.0.0
[27] HDF5Array_1.10.1 whisker_0.3-2
[29] later_0.7.5 pracma_2.2.9
[31] git2r_0.26.1 tibble_2.1.1
[33] ggplot2_3.2.1 conicfit_1.0.4
[35] lazyeval_0.2.1 magrittr_1.5
[37] crayon_1.3.4 evaluate_0.12
[39] fs_1.3.1 doParallel_1.0.14
[41] MASS_7.3-51.1 beeswarm_0.2.3
[43] geigen_2.3 tools_3.5.1
[45] scater_1.10.1 stringr_1.3.1
[47] Rhdf5lib_1.4.3 munsell_0.5.0
[49] compiler_3.5.1 rlang_0.4.0
[51] rhdf5_2.26.2 grid_3.5.1
[53] RCurl_1.95-4.11 iterators_1.0.12
[55] circular_0.4-93 igraph_1.2.2
[57] bitops_1.0-6 rmarkdown_1.10
[59] boot_1.3-20 gtable_0.2.0
[61] codetools_0.2-15 reshape2_1.4.3
[63] R6_2.4.0 gridExtra_2.3
[65] knitr_1.20 dplyr_0.8.0.1
[67] workflowr_1.6.0 rprojroot_1.3-2
[69] stringi_1.2.4 ggbeeswarm_0.6.0
[71] Rcpp_1.0.3 tidyselect_0.2.5
sessionInfo()R version 3.5.1 (2018-07-02)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Scientific Linux 7.4 (Nitrogen)
Matrix products: default
BLAS/LAPACK: /software/openblas-0.2.19-el7-x86_64/lib/libopenblas_haswellp-r0.2.19.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] peco_0.99.6 SingleCellExperiment_1.4.1
[3] SummarizedExperiment_1.12.0 DelayedArray_0.8.0
[5] BiocParallel_1.16.0 matrixStats_0.55.0
[7] Biobase_2.42.0 GenomicRanges_1.34.0
[9] GenomeInfoDb_1.18.1 IRanges_2.16.0
[11] S4Vectors_0.20.1 BiocGenerics_0.28.0
loaded via a namespace (and not attached):
[1] viridis_0.5.1 genlasso_1.4
[3] viridisLite_0.3.0 foreach_1.4.4
[5] DelayedMatrixStats_1.4.0 assertthat_0.2.1
[7] vipor_0.4.5 GenomeInfoDbData_1.2.0
[9] yaml_2.2.0 pillar_1.3.1
[11] backports_1.1.2 lattice_0.20-38
[13] glue_1.3.0 digest_0.6.20
[15] promises_1.0.1 XVector_0.22.0
[17] colorspace_1.3-2 plyr_1.8.4
[19] htmltools_0.3.6 httpuv_1.4.5
[21] Matrix_1.2-17 pkgconfig_2.0.3
[23] zlibbioc_1.28.0 purrr_0.3.2
[25] mvtnorm_1.0-11 scales_1.0.0
[27] HDF5Array_1.10.1 whisker_0.3-2
[29] later_0.7.5 pracma_2.2.9
[31] git2r_0.26.1 tibble_2.1.1
[33] ggplot2_3.2.1 conicfit_1.0.4
[35] lazyeval_0.2.1 magrittr_1.5
[37] crayon_1.3.4 evaluate_0.12
[39] fs_1.3.1 doParallel_1.0.14
[41] MASS_7.3-51.1 beeswarm_0.2.3
[43] geigen_2.3 tools_3.5.1
[45] scater_1.10.1 stringr_1.3.1
[47] Rhdf5lib_1.4.3 munsell_0.5.0
[49] compiler_3.5.1 rlang_0.4.0
[51] rhdf5_2.26.2 grid_3.5.1
[53] RCurl_1.95-4.11 iterators_1.0.12
[55] circular_0.4-93 igraph_1.2.2
[57] bitops_1.0-6 rmarkdown_1.10
[59] boot_1.3-20 gtable_0.2.0
[61] codetools_0.2-15 reshape2_1.4.3
[63] R6_2.4.0 gridExtra_2.3
[65] knitr_1.20 dplyr_0.8.0.1
[67] workflowr_1.6.0 rprojroot_1.3-2
[69] stringi_1.2.4 ggbeeswarm_0.6.0
[71] Rcpp_1.0.3 tidyselect_0.2.5