Last updated: 2018-08-09

workflowr checks: (Click a bullet for more information)

  • ✔ R Markdown file: up-to-date

    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.

  • ✔ Environment: empty

    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.

  • ✔ Seed: set.seed(12345)

    The command set.seed(12345) 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.

  • ✔ Session information: recorded

    Great job! Recording the operating system, R version, and package versions is critical for reproducibility.

  • ✔ Repository version: 6893542

    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/table-s1.txt
    Untracked:  analysis/table-s2.txt
    Untracked:  code/tb-scratch.R
    Untracked:  data/counts_per_sample.txt
    Untracked:  docs/table-s1.txt
    Untracked:  docs/table-s2.txt
    Untracked:  factorial-dox.rds

Unstaged changes:
    Modified:   data/cll-eset.rds
    Modified:   data/cll-fit2.rds
    Modified:   data/cll.RData
    Modified:   data/stem-eset.rds
    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.
Expand here to see past versions:
    File Version Author Date Message
    Rmd 6893542 John Blischak 2018-08-09 Organize analysis of batch effects in TB data.

An example of diagnosing and correcting batch effects from one of my own studies on the response to infection with Mycobacterium tuberculosis (paper, code, data).

Setup

library(dplyr)
library(limma)
library(edgeR)
# Have to load Biobase after dplyr so that exprs function works
library(Biobase)

Download data.

file_url <- "https://bitbucket.org/jdblischak/tb-data/raw/bc0f64292eb8c42a372b3a2d50e3d871c70c202e/counts_per_sample.txt"
full <- read.delim(file_url, stringsAsFactors = FALSE)

Convert to ExpressionSet.

dim(full)
[1]   156 19419
full <- full[order(full$dir), ]
rownames(full) <- paste(full$ind, full$bact, full$time, sep = ".")

x <- t(full[, grep("ENSG", colnames(full))])
p <- full %>% select(ind, bact, time, extr, rin)
stopifnot(colnames(x) == rownames(p))

eset <- ExpressionSet(assayData = x,
                      phenoData = AnnotatedDataFrame(p))

Filter lowly expressed genes.

keep <- rowSums(cpm(exprs(eset)) > 1) > 6
sum(keep)
[1] 12728
eset <- eset[keep, ]
dim(eset)
Features  Samples 
   12728      156

Normalize with TMM.

norm_factors <- calcNormFactors(exprs(eset))
exprs(eset) <- cpm(exprs(eset), lib.size = colSums(exprs(eset)) * norm_factors,
                   log = TRUE)
plotDensities(eset, legend = FALSE)

Clean up phenotype data frame to focus on early versus late timepoint for this example.

pData(eset)[, "infection"] <- ifelse(pData(eset)[, "bact"] == "none",
                                     "con", "inf")

pData(eset)[, "time"] <- ifelse(pData(eset)[, "time"] == 4,
                                "early", "late")

pData(eset)[, "batch"] <- sprintf("b%02d", pData(eset)[, "extr"])

table(pData(eset)[, c("time", "batch")])
       batch
time    b01 b02 b03 b04 b05 b06 b07 b08 b09 b10 b11 b12 b13
  early   4   4   4   4   4   4   4   4   4   4   4   4   6
  late    8   8   8   8   8   8   8   8   8   8   8   8   6

Remove batch effect

Visualize principal components 1 and 2 for the original data.

plotMDS(eset, labels = pData(eset)[, "time"], gene.selection = "common")

Remove the effect of the technical variables: batch (discrete) and RIN (continuous; a measure of RNA quality).

exprs(eset) <- removeBatchEffect(eset, batch = pData(eset)[, "batch"],
                                 covariates = pData(eset)[, "rin"])

Visualize principal components 1 and 2 for the corrected data.

plotMDS(eset, labels = pData(eset)[, "time"], gene.selection = "common")

Session information

sessionInfo()
R version 3.4.4 (2018-03-15)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 18.04.1 LTS

Matrix products: default
BLAS: /usr/lib/x86_64-linux-gnu/blas/libblas.so.3.7.1
LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.3.7.1

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  stats     graphics  grDevices utils     datasets  methods  
[8] base     

other attached packages:
[1] Biobase_2.36.2      BiocGenerics_0.22.1 edgeR_3.18.1       
[4] limma_3.32.2        dplyr_0.7.5        

loaded via a namespace (and not attached):
 [1] Rcpp_0.12.18         knitr_1.20           bindr_0.1.1         
 [4] whisker_0.3-2        magrittr_1.5         workflowr_1.1.1.9000
 [7] tidyselect_0.2.4     lattice_0.20-35      R6_2.2.2            
[10] rlang_0.2.1.9000     stringr_1.3.1        tools_3.4.4         
[13] grid_3.4.4           R.oo_1.22.0          git2r_0.23.0        
[16] htmltools_0.3.6      yaml_2.2.0           rprojroot_1.3-2     
[19] digest_0.6.15        assertthat_0.2.0     tibble_1.4.2        
[22] bindrcpp_0.2.2       purrr_0.2.5          R.utils_2.6.0       
[25] glue_1.3.0           evaluate_0.11        rmarkdown_1.10      
[28] stringi_1.2.4        pillar_1.1.0         compiler_3.4.4      
[31] backports_1.1.2-9000 R.methodsS3_1.7.1    locfit_1.5-9.1      
[34] pkgconfig_2.0.1

This reproducible R Markdown analysis was created with workflowr 1.1.1.9000