Normalization and filtering of bulkRNAseq data

workflowr

• Summary
• Checks
• Past versions

Last updated: 2021-01-21

Checks: 7 0

Knit directory: invitroOA_pilot_repository/

This reproducible R Markdown analysis was created with workflowr (version 1.6.2). The Checks tab describes the reproducibility checks that were applied when the results were created. The Past versions tab lists the development history.

---

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(20210119)

The command set.seed(20210119) 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.

Cache: none

Nice! There were no cached chunks for this analysis, so you can be confident that you successfully produced the results during this run.

File paths: relative

Great job! Using relative paths to the files within your workflowr project makes it easier to run your code on other machines.

Repository version: 78cfbcd

Great! You are using Git for version control. Tracking code development and connecting the code version to the results is critical for reproducibility.

The results in this page were generated with repository version 78cfbcd. See the Past versions tab to see a history of the changes made to the R Markdown and HTML files.

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:    .DS_Store
    Ignored:    .Rhistory
    Ignored:    .Rproj.user/
    Ignored:    data/.DS_Store
    Ignored:    data/external_scRNA/.DS_Store

Unstaged changes:
    Deleted:    analysis/figure/powerAnalysis.Rmd/curve-1.png
    Deleted:    analysis/figure/powerAnalysis.Rmd/hypoxia-1.png
    Deleted:    analysis/figure/powerAnalysis.Rmd/hypoxia-2.png
    Deleted:    analysis/figure/powerAnalysis.Rmd/hypoxia-3.png
    Deleted:    analysis/figure/powerAnalysis.Rmd/hypoxia-4.png
    Deleted:    analysis/figure/powerAnalysis.Rmd/hypoxia-5.png
    Deleted:    analysis/figure/powerAnalysis.Rmd/hypoxia-6.png
    Deleted:    analysis/figure/powerAnalysis.Rmd/hypoxia-7.png
    Deleted:    analysis/figure/powerAnalysis.Rmd/hypoxia-8.png

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 repository in which changes were made to the R Markdown (analysis/normFilter_bulkRNA.Rmd) and HTML (docs/normFilter_bulkRNA.html) files. If you’ve configured a remote Git repository (see ?wflow_git_remote), click on the hyperlinks in the table below to view the files as they were in that past version.

File	Version	Author	Date	Message
Rmd	e3fbd0f	Anthony Hung	2021-01-21	add ward et al data files to data
html	e3fbd0f	Anthony Hung	2021-01-21	add ward et al data files to data
Rmd	0711682	Anthony Hung	2021-01-21	start normalization file
html	0711682	Anthony Hung	2021-01-21	start normalization file
Rmd	28f57fa	Anthony Hung	2021-01-19	Add files for analysis

---

Introduction

This code takes the raw count data from bulk RNA sequencing and performs upper quartile normalziation on the data, filtering out lowly expressed genes. It then saves the resulting data and filtered count data (unnormalized) as R objects.

Here we also plot PCA plots to visualize the structure of the full data set (PCA plots for the data set excluding the sample that failed QC are located in the Modeling and removing unwanted technical variation page).

Load data and libraries

Here we are loading both the full data and the data without the one sample that failed QC.

library("gplots")

Attaching package: 'gplots'

The following object is masked from 'package:stats':

    lowess

library("ggplot2")
library("reshape")
library("edgeR")

Loading required package: limma

library("RColorBrewer")
library("scales")
library("cowplot")

Attaching package: 'cowplot'

The following object is masked from 'package:reshape':

    stamp

theme_set(theme_cowplot())
library("dplyr")

Attaching package: 'dplyr'

The following object is masked from 'package:reshape':

    rename

The following objects are masked from 'package:stats':

    filter, lag

The following objects are masked from 'package:base':

    intersect, setdiff, setequal, union

# Load colors 
colors <- colorRampPalette(c(brewer.pal(9, "Blues")[1],brewer.pal(9, "Blues")[9]))(100)
pal <- c(brewer.pal(9, "Set1"), brewer.pal(8, "Set2"), brewer.pal(12, "Set3"))


# load in relabeled counts
raw_counts <- readRDS("data/raw_counts_relabeled.rds")
raw_counts_includeNA19160 <- readRDS("data/raw_counts_relabeled_includeNA19160.rds")

# Create DGEList object to allow for easier application of different normalization methods
raw_counts <- DGEList(raw_counts, group = colnames(raw_counts))
raw_counts_includeNA19160 <- DGEList(raw_counts_includeNA19160, group = colnames(raw_counts_includeNA19160))

# load in reordered sample information
sampleinfo <- readRDS("data/Sample.info.RNAseq.reordered.rds")
sampleinfo_includeNA19160 <- readRDS("data/Sample.info.RNAseq.reordered_includeNA19160.rds")

Upperquartile Normalization

Apply upperquartile normalization to the data.

upperquartile <- calcNormFactors(raw_counts, method = "upperquartile")
upperquartile <- cpm(upperquartile, log=TRUE, normalized.lib.sizes = T)
head(upperquartile)

                18855_3_S 19160_3_S 18856_3_U 18856_1_U 18855_2_S 18856_2_S
ENSG00000223972 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986
ENSG00000227232 -1.376624 -2.046788 -2.154778 -2.585632 -2.545595 -1.167335
ENSG00000278267 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986
ENSG00000243485 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986
ENSG00000284332 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986
ENSG00000237613 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986
                19160_3_U 18855_2_U 19160_2_S 18855_1_S 18856_1_S 19160_1_S
ENSG00000223972 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986
ENSG00000227232 -3.129986 -2.527469 -1.317333 -1.415729 -3.129986 -2.313104
ENSG00000278267 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986
ENSG00000243485 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986
ENSG00000284332 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986
ENSG00000237613 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986
                19160_2_U 18855_1_U 18856_3_S 18856_2_U 18855_3_U
ENSG00000223972 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986
ENSG00000227232 -2.601918 -1.247154 -2.084127 -1.609909 -3.129986
ENSG00000278267 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986
ENSG00000243485 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986
ENSG00000284332 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986
ENSG00000237613 -3.129986 -3.129986 -3.129986 -3.129986 -3.129986

strained <- sampleinfo$treatment == "Strain"
unstrained <- sampleinfo$treatment == "Unstrain"
ind_1 <- sampleinfo$Individual == "NA18855 "
ind_2 <- sampleinfo$Individual == "NA18856 "
ind_3 <- sampleinfo$Individual == "NA19160 "


# Look at density plots for all individuals broken down by each treatment type
col = as.data.frame(pal[as.numeric(sampleinfo$Individual)])

plotDensities(upperquartile[,strained], col=col[strained, ], legend="topright")

Past versions of “upperquartile normalization-1.png”

Version	Author	Date
0711682	Anthony Hung	2021-01-21

plotDensities(upperquartile[,unstrained], col=col[unstrained, ], legend="topright")

Past versions of “upperquartile normalization-2.png”

Version	Author	Date
0711682	Anthony Hung	2021-01-21

# Look at density plots broken down by individual
col = as.data.frame(pal[as.numeric(sampleinfo$treatment)])

plotDensities(upperquartile[,ind_1], col=col[ind_1, ], legend="topright")

Past versions of “upperquartile normalization-3.png”

Version	Author	Date
0711682	Anthony Hung	2021-01-21

plotDensities(upperquartile[,ind_2], col=col[ind_2, ], legend="topright")

Past versions of “upperquartile normalization-4.png”

Version	Author	Date
0711682	Anthony Hung	2021-01-21

plotDensities(upperquartile[,ind_3], col=col[ind_3, ], legend="topright")

Past versions of “upperquartile normalization-5.png”

Version	Author	Date
0711682	Anthony Hung	2021-01-21

Boxplots of upperquartile across samples

meltupperquartile <- melt(upperquartile)
names(meltupperquartile) <- c("gene", "sampleID", "upperquartile")
p <- ggplot(meltupperquartile, aes(factor(sampleID), upperquartile)) 
p + geom_boxplot() + theme(axis.text.x = element_text(angle = 90))

Past versions of “boxplot upperquartile-1.png”

Version	Author	Date
0711682	Anthony Hung	2021-01-21

Filtering for lowly expressed genes

The criteria we are using is an average log2upperquartile > 2.5 in at least 4 samples. After this filtering step, 10486 genes remain.

cutoff <- 2.5

keep <- rowSums( upperquartile > cutoff ) >=4

counts_upperquartile <- raw_counts[keep,]
filtered_upperquartile <- upperquartile[keep,]
dim(filtered_upperquartile)

[1] 10486    17

Boxplots of normalized+filtered counts across samples

After normalization, the data do indeed look more normal than when unnormalized.

melt_filt_upperquartile <- melt(filtered_upperquartile)
names(melt_filt_upperquartile) <- c("gene", "sampleID", "log2upperquartile")
p1 <- ggplot(melt_filt_upperquartile, aes(factor(sampleID), log2upperquartile)) 
p1 + geom_boxplot() + theme(axis.text.x = element_text(angle = 90))

Past versions of “boxplot filt upperquartile-1.png”

Version	Author	Date
0711682	Anthony Hung	2021-01-21

plotDensities(filtered_upperquartile, legend = F)

Past versions of “boxplot filt upperquartile-2.png”

Version	Author	Date
0711682	Anthony Hung	2021-01-21

Save normalized/filtered count matrix and filtered count matrix

saveRDS(counts_upperquartile, "data/filtered_counts.rds")
saveRDS(filtered_upperquartile, "data/norm_filtered_counts.rds")

PCA plots for dataset including the sample filtered out by QC

upperquartile Normalization of full data

upperquartile_includeNA19160 <- calcNormFactors(raw_counts_includeNA19160, method = "upperquartile")
upperquartile_includeNA19160 <- cpm(upperquartile_includeNA19160, log=TRUE, normalized.lib.sizes = T)
head(upperquartile_includeNA19160)

                18855_3_S 19160_3_S 18856_3_U 18856_1_U 18855_2_S 18856_2_S
ENSG00000223972 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145
ENSG00000227232 -1.371277 -2.027812 -2.134933 -2.565487 -2.526602 -1.157511
ENSG00000278267 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145
ENSG00000243485 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145
ENSG00000284332 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145
ENSG00000237613 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145
                19160_3_U 18855_2_U 19160_2_S 18855_1_S 18856_1_S 19160_1_S
ENSG00000223972 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145
ENSG00000227232 -3.101145 -2.508875 -1.294699 -1.401586 -3.101145 -2.296780
ENSG00000278267 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145
ENSG00000243485 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145
ENSG00000284332 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145
ENSG00000237613 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145
                19160_2_U 19160_1_U 18855_1_U 18856_3_S 18856_2_U 18855_3_U
ENSG00000223972 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145
ENSG00000227232 -2.577970 -2.336134 -1.239235 -2.062052 -1.601659 -3.101145
ENSG00000278267 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145
ENSG00000243485 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145
ENSG00000284332 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145
ENSG00000237613 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145 -3.101145

strained <- sampleinfo_includeNA19160$treatment == "Strain"
unstrained <- sampleinfo_includeNA19160$treatment == "Unstrain"
ind_1 <- sampleinfo_includeNA19160$Individual == "NA18855 "
ind_2 <- sampleinfo_includeNA19160$Individual == "NA18856 "
ind_3 <- sampleinfo_includeNA19160$Individual == "NA19160 "


# Look at density plots for all individuals broken down by each treatment type
col = as.data.frame(pal[as.numeric(sampleinfo_includeNA19160$Individual)])

plotDensities(upperquartile_includeNA19160[,strained], col=col[strained, ], legend="topright")

Past versions of “upperquartile normalization full data-1.png”

Version	Author	Date
e3fbd0f	Anthony Hung	2021-01-21

plotDensities(upperquartile_includeNA19160[,unstrained], col=col[unstrained, ], legend="topright")

Past versions of “upperquartile normalization full data-2.png”

Version	Author	Date
e3fbd0f	Anthony Hung	2021-01-21

# Look at density plots broken down by individual
col = as.data.frame(pal[as.numeric(sampleinfo_includeNA19160$treatment)])

plotDensities(upperquartile_includeNA19160[,ind_1], col=col[ind_1, ], legend="topright")

Past versions of “upperquartile normalization full data-3.png”

Version	Author	Date
e3fbd0f	Anthony Hung	2021-01-21

plotDensities(upperquartile_includeNA19160[,ind_2], col=col[ind_2, ], legend="topright")

Past versions of “upperquartile normalization full data-4.png”

Version	Author	Date
e3fbd0f	Anthony Hung	2021-01-21

plotDensities(upperquartile_includeNA19160[,ind_3], col=col[ind_3, ], legend="topright")

Past versions of “upperquartile normalization full data-5.png”

Version	Author	Date
e3fbd0f	Anthony Hung	2021-01-21

Boxplots of upperquartile across samples

meltupperquartile_includeNA19160 <- melt(upperquartile_includeNA19160)
names(meltupperquartile_includeNA19160) <- c("gene", "sampleID", "upperquartile")
p <- ggplot(meltupperquartile_includeNA19160, aes(factor(sampleID), upperquartile_includeNA19160)) 
p + geom_boxplot() + theme(axis.text.x = element_text(angle = 90))

Past versions of “boxplot upperquartile full data-1.png”

Version	Author	Date
e3fbd0f	Anthony Hung	2021-01-21

Filtering for lowly expressed genes (avg log2upperquartile > 2.5 in at least 4 samples)

cutoff <- 2.5

keep <- rowSums( upperquartile_includeNA19160 > cutoff ) >=4

counts_upperquartile_includeNA19160 <- raw_counts_includeNA19160[keep,]
filtered_upperquartile_includeNA19160 <- upperquartile_includeNA19160[keep,]
dim(filtered_upperquartile_includeNA19160)

[1] 10517    18

Boxplots of normalized+filtered counts across samples

melt_filt_upperquartile_includeNA19160 <- melt(filtered_upperquartile_includeNA19160)
names(melt_filt_upperquartile_includeNA19160) <- c("gene", "sampleID", "log2upperquartile")
p1 <- ggplot(melt_filt_upperquartile_includeNA19160, aes(factor(sampleID), log2upperquartile)) 
p1 + geom_boxplot() + theme(axis.text.x = element_text(angle = 90))

Past versions of “boxplot filt upperquartile full data-1.png”

Version	Author	Date
e3fbd0f	Anthony Hung	2021-01-21

plotDensities(filtered_upperquartile_includeNA19160, legend = F)

Past versions of “boxplot filt upperquartile full data-2.png”

Version	Author	Date
e3fbd0f	Anthony Hung	2021-01-21

Plot PCA

Reassuringly, the sample that was removed for low numbers of mapped reads still clusters as expected with its technical replicates.

#Load PCA plotting Function
source("code/PCA_fn.R")

# Perform PCA
pca_genes <- prcomp(t(filtered_upperquartile_includeNA19160), scale = T)
scores <- pca_genes$x

variances <- pca_genes$sdev^2
explained <- variances / sum(variances)
plot(pca_genes, main = "Variance per PC")

Past versions of “PCA full data-1.png”

Version	Author	Date
e3fbd0f	Anthony Hung	2021-01-21

#Make PCA plots with the factors colored by Individual

### PCA norm+filt Data
for (n in 1:5){
  col.v <- pal[as.integer(sampleinfo_includeNA19160$Individual)]
  plot_scores(pca_genes, scores, n, n+1, col.v)
}

Past versions of “PCA full data-2.png”

Version	Author	Date
e3fbd0f	Anthony Hung	2021-01-21

Past versions of “PCA full data-3.png”

Version	Author	Date
e3fbd0f	Anthony Hung	2021-01-21

Past versions of “PCA full data-4.png”

Version	Author	Date
e3fbd0f	Anthony Hung	2021-01-21

Past versions of “PCA full data-5.png”

Version	Author	Date
e3fbd0f	Anthony Hung	2021-01-21

Past versions of “PCA full data-6.png”

Version	Author	Date
e3fbd0f	Anthony Hung	2021-01-21

#make PCA plots with symbols as treatment status and colors as individuals for figure
library(ggfortify)
autoplot(pca_genes, data = sampleinfo_includeNA19160, colour = "Individual", shape = "treatment", size = 3) + 
     theme_cowplot() +
     theme(legend.position = "none")

Past versions of “PCA full data-7.png”

Version	Author	Date
e3fbd0f	Anthony Hung	2021-01-21

autoplot(pca_genes, data = sampleinfo_includeNA19160, colour = "Individual", shape = "treatment") + 
     theme_cowplot()

Past versions of “PCA full data-8.png”

Version	Author	Date
e3fbd0f	Anthony Hung	2021-01-21

autoplot(pca_genes, data = sampleinfo_includeNA19160, colour = "Individual", shape = "treatment", size = 3, x = 3, y = 4) + 
     theme_cowplot() +
     theme(legend.position = "none")

Past versions of “PCA full data-9.png”

Version	Author	Date
e3fbd0f	Anthony Hung	2021-01-21

autoplot(pca_genes, data = sampleinfo_includeNA19160, colour = "Individual", shape = "treatment", size = 3, x = 5, y = 6) + 
     theme_cowplot() +
     theme(legend.position = "none")

Past versions of “PCA full data-10.png”

Version	Author	Date
e3fbd0f	Anthony Hung	2021-01-21

Session information

sessionInfo()

R version 3.6.3 (2020-02-29)
Platform: x86_64-apple-darwin15.6.0 (64-bit)
Running under: macOS Catalina 10.15.7

Matrix products: default
BLAS:   /Library/Frameworks/R.framework/Versions/3.6/Resources/lib/libRblas.0.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/3.6/Resources/lib/libRlapack.dylib

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] ggfortify_0.4.10   dplyr_1.0.2        cowplot_1.1.1      scales_1.1.1      
 [5] RColorBrewer_1.1-2 edgeR_3.26.8       limma_3.40.6       reshape_0.8.8     
 [9] ggplot2_3.3.2      gplots_3.0.4      

loaded via a namespace (and not attached):
 [1] gtools_3.8.2           tidyselect_1.1.0       locfit_1.5-9.4        
 [4] xfun_0.16              purrr_0.3.4            lattice_0.20-41       
 [7] colorspace_1.4-1       vctrs_0.3.4            generics_0.0.2        
[10] htmltools_0.5.0        yaml_2.2.1             rlang_0.4.7           
[13] later_1.1.0.1          pillar_1.4.6           glue_1.4.2            
[16] withr_2.2.0            lifecycle_0.2.0        plyr_1.8.6            
[19] stringr_1.4.0          munsell_0.5.0          gtable_0.3.0          
[22] workflowr_1.6.2        caTools_1.18.0         evaluate_0.14         
[25] labeling_0.3           knitr_1.29             httpuv_1.5.4          
[28] Rcpp_1.0.5             KernSmooth_2.23-17     promises_1.1.1        
[31] backports_1.1.9        gdata_2.18.0           farver_2.0.3          
[34] fs_1.5.0               gridExtra_2.3          digest_0.6.25         
[37] stringi_1.4.6          grid_3.6.3             rprojroot_1.3-2       
[40] tools_3.6.3            bitops_1.0-6           magrittr_1.5          
[43] tibble_3.0.3           tidyr_1.1.2            crayon_1.3.4          
[46] whisker_0.4            pkgconfig_2.0.3        ellipsis_0.3.1        
[49] rmarkdown_2.3          rstudioapi_0.11.0-9000 R6_2.4.1              
[52] git2r_0.27.1           compiler_3.6.3