Last updated: 2020-03-31
Checks: 7 0
Knit directory: misc/
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(20191122) 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/contrainedclustering.Rmd
Untracked: analysis/gsea.Rmd
Untracked: analysis/ideas.Rmd
Untracked: analysis/methylation.Rmd
Untracked: analysis/susie.Rmd
Untracked: code/sccytokines.R
Untracked: code/scdeCalibration.R
Untracked: data/bart/
Untracked: data/cytokine/DE_controls_output_filter10.RData
Untracked: data/cytokine/DE_controls_output_filter10_addlimma.RData
Untracked: data/cytokine/README
Untracked: data/cytokine/test.RData
Untracked: data/cytokine_normalized.RData
Untracked: data/deconv/
Untracked: data/scde/
Unstaged changes:
Modified: analysis/binomthinultimate.Rmd
Modified: analysis/deconvolution.Rmd
Modified: analysis/index.Rmd
Modified: analysis/limma.Rmd
Deleted: data/mout_high.RData
Deleted: data/scCDT.RData
Deleted: data/sva_sva_high.RData
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 | 4f066db | DongyueXie | 2020-03-31 | wflow_publish(“analysis/scdeCheckAlpha.Rmd”) |
| html | c0fdf5c | DongyueXie | 2020-03-19 | Build site. |
| Rmd | a7a0098 | DongyueXie | 2020-03-19 | wflow_publish(“analysis/scdeCheckAlpha.Rmd”) |
Compare the loglikelihood from sva-limma-ash, setting \(\alpha=0\) and \(\alpha=1\), on real data.
Even if we do not take log, \(\alpha=1\) still give larger loglikelihood.
library(vicar)
library(sva)
library(cate)
library(seqgendiff)
library(edgeR)
library(MASS)
load('data/scde/scCD4.RData')
load('data/scde/scCD8.RData')
load('data/scde/scCD14.RData')
load('data/scde/scMB.RData')
mglmfit = function(Y,X){
G = nrow(Y)
coefs = c()
ses = c()
for(i in 1:G){
fit = glm(y~.,data.frame(y=Y[i,],x=X[,-1]),family = 'poisson')
coefs = rbind(coefs,fit$coefficients)
ses = rbind(ses,summary(fit)$coefficients[,2])
}
return(list(coefficients=coefs,se = ses))
}CD4, CD8, CD14 and B cells. Look at number of genes and cells.
dim(CD4)[1] 13713 709dim(CD8)[1] 13713 313dim(CD14)[1] 13713 432dim(MB)[1] 13713 342Y = as.matrix(cbind(CD4,MB))
group_idx = c(rep(1,dim(CD4)[2]),rep(0,dim(MB)[2]))
# remove genes appearing in less than 10 cells
Y = Y[-which(rowSums(Y!=0)<10),]
X = model.matrix(~group_idx)
real_study = function(Y,X,run_sva_eb=TRUE,n.sv=NULL){
if(run_sva_eb){
sva_sva = sva((Y),mod=X,mod0=X[,1],n.sv = n.sv)
X.sva <- cbind(X, sva_sva$sv)
lmout <- limma::lmFit(object = (Y), design = X.sva)
eout <- limma::eBayes(lmout)
svaout <- list()
svaout$betahat <- lmout$coefficients[, 2]
svaout$sebetahat <- lmout$stdev.unscaled[, 2] * sqrt(eout$s2.post)
svaout$pvalues <- eout$p.value[, 2]
sva_limma_ash0 = ashr::ash(svaout$betahat,svaout$sebetahat,alpha=0)
sva_limma_ash1 = ashr::ash(svaout$betahat,svaout$sebetahat,alpha=1)
print(paste('sva-limma-ash0 loglik: ',round(sva_limma_ash0$loglik,2),';',
'sva-limma-ash1 loglik: ',round(sva_limma_ash1$loglik)))
}else{
lmout <- limma::lmFit(object = (Y), design = X)
out = list()
out$betahat <- lmout$coefficients[, 2]
out$sebetahat <- lmout$stdev.unscaled[, 2] * lmout$sigma
lm_ash0 = ashr::ash(out$betahat,out$sebetahat,alpha=0)
lm_ash1 = ashr::ash(out$betahat,out$sebetahat,alpha=1)
print(paste('lm-ash0 loglik: ',round(lm_ash0$loglik,2),';',
'lm-ash1 loglik: ',round(lm_ash1$loglik)))
}
}
real_study(Y,X,3)Number of significant surrogate variables is: 3
Iteration (out of 5 ):1 2 3 4 5 [1] "sva-limma-ash0 loglik: 16134.95 ; sva-limma-ash1 loglik: 17053"real_study(Y,X,run_sva_eb = FALSE)[1] "lm-ash0 loglik: 16049.71 ; lm-ash1 loglik: 17018"Take log1p of Y
real_study(log(Y+0.5),X,14)Number of significant surrogate variables is: 14
Iteration (out of 5 ):1 2 3 4 5 [1] "sva-limma-ash0 loglik: 16929.86 ; sva-limma-ash1 loglik: 17545"real_study(log(Y+0.5),X,run_sva_eb = FALSE)[1] "lm-ash0 loglik: 16628.98 ; lm-ash1 loglik: 17339"Voom-lm-ash
d = DGEList(Y)
d = calcNormFactors(d)
y = voom(d, design=X, plot = FALSE)
real_study(y,X,run_sva_eb = FALSE)[1] "lm-ash0 loglik: 2165.18 ; lm-ash1 loglik: 1471"glm-ash
fit = mglmfit(Y,X)
glm_ash0 = ashr::ash(fit$coefficients[,2],fit$se[,2],alpha=0)
glm_ash1 = ashr::ash(fit$coefficients[,2],fit$se[,2],alpha=1)
print(paste('glm-ash0 loglik: ',round(glm_ash0$loglik,2),';',
'glm-ash1 loglik: ',round(glm_ash1$loglik)))[1] "glm-ash0 loglik: -9342.16 ; glm-ash1 loglik: -9895"Y = as.matrix(cbind(CD8,CD14))
group_idx = c(rep(1,dim(CD8)[2]),rep(0,dim(CD14)[2]))
# remove genes appearing in less than 10 cells
Y = Y[-which(rowSums(Y!=0)<10),]
X = model.matrix(~group_idx)
real_study(Y,X,3)Number of significant surrogate variables is: 3
Iteration (out of 5 ):1 2 3 4 5 [1] "sva-limma-ash0 loglik: 14146.63 ; sva-limma-ash1 loglik: 15014"real_study(Y,X,run_sva_eb = FALSE)[1] "lm-ash0 loglik: 13372.89 ; lm-ash1 loglik: 14499"Take log1p of Y
real_study(log(Y+0.5),X,27)Number of significant surrogate variables is: 27
Iteration (out of 5 ):1 2 3 4 5 [1] "sva-limma-ash0 loglik: 16140.94 ; sva-limma-ash1 loglik: 16406"real_study(log(Y+0.5),X,run_sva_eb = FALSE)[1] "lm-ash0 loglik: 14028.98 ; lm-ash1 loglik: 14877"Voom-lm-ash
d = DGEList(Y)
d = calcNormFactors(d)
y = voom(d, design=X, plot = FALSE)
real_study(y,X,run_sva_eb = FALSE)[1] "lm-ash0 loglik: 117.68 ; lm-ash1 loglik: -1781"glm-ash
fit = mglmfit(Y,X)
glm_ash0 = ashr::ash(fit$coefficients[,2],fit$se[,2],alpha=0)
glm_ash1 = ashr::ash(fit$coefficients[,2],fit$se[,2],alpha=1)
print(paste('glm-ash0 loglik: ',round(glm_ash0$loglik,2),';',
'glm-ash1 loglik: ',round(glm_ash1$loglik)))[1] "glm-ash0 loglik: -7016.82 ; glm-ash1 loglik: -8182"Taken from here
load('data/scde/MouseJaitinSpleen.rda')
suppressPackageStartupMessages(library(SummarizedExperiment))
suppressPackageStartupMessages(library(MultiAssayExperiment))
mat = exprs(MouseJaitinSpleen)
idx = which(MouseJaitinSpleen$ERCC_dilution=='2.50E-05')
cell_type = MouseJaitinSpleen$group_name[idx]
table(cell_type)cell_type
B cell CD11c+ CD11c+(2hr_LPS)
48 2112 1536
CD8-CD4+ESAM+ CD8-pDC CD8+CD86-
96 96 96
CD8+pDC GC B cell monocyte_or_neutrophil
96 48 48
NK_cell splenocyte
48 48 mat = mat[,idx]Y = as.matrix(cbind(mat[,which(cell_type=='B cell')],mat[,which(cell_type=='NK_cell')]))
group_idx = c(rep(1,dim(mat[,which(cell_type=='B cell')])[2]),
rep(0,dim(mat[,which(cell_type=='NK_cell')])[2]))
# remove genes appearing in less than 10 cells
Y = Y[-which(rowSums(Y!=0)<10),]
rm.idx = which(colSums(Y)==0)
Y = Y[,-rm.idx]
group_idx = group_idx[-rm.idx]
X = model.matrix(~group_idx)
real_study(Y,X)Number of significant surrogate variables is: 2
Iteration (out of 5 ):1 2 3 4 5 [1] "sva-limma-ash0 loglik: -372.28 ; sva-limma-ash1 loglik: -106"real_study(Y,X,run_sva_eb = FALSE)[1] "lm-ash0 loglik: -388.83 ; lm-ash1 loglik: -120"# How about just run linear model and then ash?
X = model.matrix(~group_idx)
lmout <- limma::lmFit(object = (Y), design = X)
svaout <- list()
svaout$betahat <- lmout$coefficients[, 2]
svaout$sebetahat <- lmout$stdev.unscaled[, 2] * lmout$sigma
svaout$pvalues <- eout$p.value[, 2]
# sva-limma-ash-alpha 0
sva_limma_ash0 = ashr::ash(svaout$betahat,svaout$sebetahat,alpha=0)
sva_limma_ash0$loglik
# sva-limma-ash-alpha 1
sva_limma_ash1 = ashr::ash(svaout$betahat,svaout$sebetahat,alpha=1)
sva_limma_ash1$loglikTake log1p of y
real_study(log(Y+0.5),X)Number of significant surrogate variables is: 15
Iteration (out of 5 ):1 2 3 4 5 [1] "sva-limma-ash0 loglik: 304.17 ; sva-limma-ash1 loglik: 372"real_study(log(Y+0.5),X,run_sva_eb = FALSE)[1] "lm-ash0 loglik: -147.77 ; lm-ash1 loglik: -24"Voom-lm-ash
d = DGEList(Y)
d = calcNormFactors(d)
y = voom(d, design=X, plot = FALSE)
real_study(y,X,run_sva_eb = FALSE)[1] "lm-ash0 loglik: -216.76 ; lm-ash1 loglik: -173"glm-ash
fit = mglmfit(Y,X)
glm_ash0 = ashr::ash(fit$coefficients[,2],fit$se[,2],alpha=0)
glm_ash1 = ashr::ash(fit$coefficients[,2],fit$se[,2],alpha=1)
print(paste('glm-ash0 loglik: ',round(glm_ash0$loglik,2),';',
'glm-ash1 loglik: ',round(glm_ash1$loglik)))[1] "glm-ash0 loglik: -2043.3 ; glm-ash1 loglik: -2080"Y = as.matrix(cbind(mat[,which(cell_type=='CD8-pDC')],mat[,which(cell_type=='splenocyte')]))
group_idx = c(rep(1,dim(mat[,which(cell_type=='CD8-pDC')])[2]),
rep(0,dim(mat[,which(cell_type=='splenocyte')])[2]))
# remove genes appearing in less than 10 cells
Y = Y[-which(rowSums(Y!=0)<10),]
rm.idx = which(colSums(Y)==0)
Y = Y[,-rm.idx]
group_idx = group_idx[-rm.idx]
X = model.matrix(~group_idx)
real_study(Y,X)Number of significant surrogate variables is: 3
Iteration (out of 5 ):1 2 3 4 5 [1] "sva-limma-ash0 loglik: 186.59 ; sva-limma-ash1 loglik: 539"real_study(Y,X,run_sva_eb = FALSE)[1] "lm-ash0 loglik: 97.49 ; lm-ash1 loglik: 488"Take log1p of y
real_study(log(Y+0.5),X)Number of significant surrogate variables is: 24
Iteration (out of 5 ):1 2 3 4 5 [1] "sva-limma-ash0 loglik: 927.11 ; sva-limma-ash1 loglik: 1011"real_study(log(Y+0.5),X,run_sva_eb = FALSE)[1] "lm-ash0 loglik: 301.85 ; lm-ash1 loglik: 519"Voom-lm-ash
d = DGEList(Y)
d = calcNormFactors(d)
y = voom(d, design=X, plot = FALSE)
real_study(y,X,run_sva_eb = FALSE)[1] "lm-ash0 loglik: -122.92 ; lm-ash1 loglik: -12"glm-ash
fit = mglmfit(Y,X)
glm_ash0 = ashr::ash(fit$coefficients[,2],fit$se[,2],alpha=0)
glm_ash1 = ashr::ash(fit$coefficients[,2],fit$se[,2],alpha=1)
print(paste('glm-ash0 loglik: ',round(glm_ash0$loglik,2),';',
'glm-ash1 loglik: ',round(glm_ash1$loglik)))[1] "glm-ash0 loglik: -3605.35 ; glm-ash1 loglik: -3773"Taken here
datax=readRDS("data/scde/GSE81076-GPL18573.rds")
datax_gene = experiments(datax)[["gene"]]
mat = (assays(datax_gene)[["count"]])
cell63_idx = 1:96
TGFBR3_idx = 673:768
Y = cbind(mat[,cell63_idx],mat[,TGFBR3_idx])
group_idx = c(rep(1,length(cell63_idx)),
rep(0,length(TGFBR3_idx)))
Y = Y[-which(rowSums(Y!=0)<10),]
X = model.matrix(~group_idx)
real_study(Y,X)Number of significant surrogate variables is: 2
Iteration (out of 5 ):1 2 3 4 5 [1] "sva-limma-ash0 loglik: -5114.92 ; sva-limma-ash1 loglik: -2264"real_study(Y,X,run_sva_eb = FALSE)[1] "lm-ash0 loglik: -5100.36 ; lm-ash1 loglik: -2259"Take log1p of y
real_study(log(Y+0.5),X)Number of significant surrogate variables is: 30
Iteration (out of 5 ):1 2 3 4 5 [1] "sva-limma-ash0 loglik: 2766.62 ; sva-limma-ash1 loglik: 3803"real_study(log(Y+0.5),X,run_sva_eb = FALSE)[1] "lm-ash0 loglik: 160.62 ; lm-ash1 loglik: 1706"Voom-lm-ash
d = DGEList(Y)
d = calcNormFactors(d)
y = voom(d, design=X, plot = FALSE)
real_study(y,X,run_sva_eb = FALSE)[1] "lm-ash0 loglik: -7375.07 ; lm-ash1 loglik: -7302"glm-ash
fit = mglmfit(Y,X)
glm_ash0 = ashr::ash(fit$coefficients[,2],fit$se[,2],alpha=0)
glm_ash1 = ashr::ash(fit$coefficients[,2],fit$se[,2],alpha=1)
print(paste('glm-ash0 loglik: ',round(glm_ash0$loglik,2),';',
'glm-ash1 loglik: ',round(glm_ash1$loglik)))[1] "glm-ash0 loglik: -15292.75 ; glm-ash1 loglik: -16158"
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] MultiAssayExperiment_1.8.3 SummarizedExperiment_1.12.0
[3] DelayedArray_0.8.0 matrixStats_0.54.0
[5] Biobase_2.42.0 GenomicRanges_1.34.0
[7] GenomeInfoDb_1.18.1 IRanges_2.16.0
[9] S4Vectors_0.20.1 BiocGenerics_0.28.0
[11] MASS_7.3-51.1 edgeR_3.24.0
[13] limma_3.38.2 seqgendiff_1.2.1
[15] cate_1.0.4 sva_3.30.0
[17] BiocParallel_1.16.0 genefilter_1.64.0
[19] mgcv_1.8-25 nlme_3.1-137
[21] vicar_0.1-10
loaded via a namespace (and not attached):
[1] svd_0.4.1 bitops_1.0-6 fs_1.3.1
[4] bit64_0.9-7 doParallel_1.0.14 rprojroot_1.3-2
[7] tools_3.5.1 backports_1.1.2 R6_2.3.0
[10] DBI_1.0.0 lazyeval_0.2.1 colorspace_1.3-2
[13] tidyselect_0.2.5 gridExtra_2.3 bit_1.1-14
[16] compiler_3.5.1 git2r_0.26.1 scales_1.0.0
[19] SQUAREM_2017.10-1 mixsqp_0.2-2 stringr_1.3.1
[22] esaBcv_1.2.1 digest_0.6.18 rmarkdown_1.10
[25] XVector_0.22.0 pscl_1.5.2 pkgconfig_2.0.2
[28] htmltools_0.3.6 ruv_0.9.7 rlang_0.4.0
[31] RSQLite_2.1.1 dplyr_0.8.0.1 leapp_1.2
[34] RCurl_1.95-4.11 magrittr_1.5 GenomeInfoDbData_1.2.0
[37] Matrix_1.2-15 Rcpp_1.0.2 munsell_0.5.0
[40] stringi_1.2.4 whisker_0.3-2 yaml_2.2.0
[43] zlibbioc_1.28.0 plyr_1.8.4 grid_3.5.1
[46] blob_1.1.1 promises_1.0.1 crayon_1.3.4
[49] lattice_0.20-38 splines_3.5.1 annotate_1.60.0
[52] locfit_1.5-9.1 knitr_1.20 pillar_1.3.1
[55] corpcor_1.6.9 codetools_0.2-15 XML_3.98-1.16
[58] glue_1.3.0 evaluate_0.12 httpuv_1.4.5
[61] foreach_1.4.4 gtable_0.2.0 purrr_0.3.2
[64] assertthat_0.2.0 ashr_2.2-39 ggplot2_3.1.1
[67] xtable_1.8-3 later_0.7.5 survival_2.43-1
[70] truncnorm_1.0-8 tibble_2.1.1 iterators_1.0.10
[73] AnnotationDbi_1.44.0 memoise_1.1.0 workflowr_1.6.0