The Snf2 dataset

The RNA-Seq dataset we will use in this practical has been produced by Gierliński et al ([@pmid26206307, @pmid27022035]). The dataset is composed of 48 WT yeast samples vs 48 Snf2 knock-out mutant cell line. The prepared RNA-Seq libraries (unstranded) were pooled and sequenced on seven lanes of a single flow-cell on an Illumina HiSeq 2000 resulting in a total of 1 billion 50-bp single-end reads across the 96 samples. RNA-Seq reads have been cleaned, mapped and counted to generated a count data matrix containing 7126 rows/genes. The primary objective of this study was to check whether the observed gene read counts distribution where consistent with theorical models (e.g. negative binomial). More information can be obtained in the original paper (pdf)

Loading the dataset

R enables to download data directly from the Web. The expression matrix and phenotypic information will be loaded into R using the read.table function. Both table will be converted into a data.frame object when loaded into R. The ‘count.table’ object will contains counts for each gene (row) and each sample (column).

# Download data files from the Web site (only if not done yet)
url.counts <- "http://jvanheld.github.io/stats_avec_RStudio_EBA/practicals/yeast_2x48_replicates/data/"

## Local paths: create a directory to store the results
dir.snf2 <- ("~/ASG/practicals/rnaseq_snf2_Schurch_2015")
dir.counts <- file.path(dir.snf2, "data")
file.counts <- file.path(dir.counts, "counts.txt")
file.expDesign <- file.path(dir.counts, "expDesign.txt")

## Create a directory to download the dataset if it does not exist
dir.create(dir.counts, showWarnings = FALSE, recursive = TRUE)

## Download the data files if required
if (!file.exists(file.counts)) {
  message("Downloading count table from ", url.counts)
  download.file(url=file.path(url.counts, "counts.txt"), destfile = file.counts)
}
if (!file.exists(file.expDesign)) {
  message("Downloading design table from ", url.counts)
  download.file(url=file.path(url.counts, "expDesign.txt"), destfile = file.expDesign)
}

# Load the count table
count.table <- read.table(file=file.counts, sep="\t", header=TRUE, row.names=1)
# View(count.table)

Phenotypic data

The dataset contains RNA-Seq count data for a wild type strain (WT) and a Snf2 mutant, with 48 biological replicates for each genotype.

All phenotypic informations are enclosed in a dedicated file. Note that the produced data.frame encodes the ‘strains’ columns as a factor1.

# Load experimental design file
expDesign <- read.table(file=file.expDesign, sep="\t", header=TRUE)
#View(expDesign)

# Check the first and last line of the phenotypic data
head(expDesign)
  label strain
1   WT1     WT
2   WT2     WT
3   WT3     WT
4   WT4     WT
5   WT5     WT
6   WT6     WT
tail(expDesign)
   label strain
91 Snf43    Snf
92 Snf44    Snf
93 Snf45    Snf
94 Snf46    Snf
95 Snf47    Snf
96 Snf48    Snf
## Count the number of sample in each class
table(expDesign$strain)

Snf  WT 
 48  48 
## Define a strain-specific color for each sample,
## and add it as a supplementary column to the phenotypic data
col.strain <- c(WT="green", Snf="orange") # Choose one color per strain
expDesign$color <- col.strain[as.vector(expDesign$strain)]
  • Draw a barplot showing the number of reads in each sample. Use either the colSums() or the apply() function (run help(colSums() if you don’t know this function).
  • What can you say from this diagram?

View solution| Hide solution

Descriptive statistics

Basic statistics

Before going further in the analysis, we will compute some descriptive statistics on the dataset. At this stage we only compute statistics per sample, since statistics per gene are meaningful only after library-wise normalization of the counts.

## Dimensions
nb.samples <- ncol(count.table)
print(nb.samples)
[1] 96
nb_genes <- nrow(count.table)
print(nb_genes)
[1] 7126
dim(count.table)
[1] 7126   96
## Min, Max, median (...). 
## Here on the first 4 samples
head(summary(count.table[,1:4]))
      WT1              WT2              WT3              WT4        
 Min.   :     0   Min.   :     0   Min.   :     0   Min.   :     0  
 1st Qu.:    49   1st Qu.:    88   1st Qu.:    74   1st Qu.:   104  
 Median :   224   Median :   371   Median :   297   Median :   430  
 Mean   :   837   Mean   :  1107   Mean   :   903   Mean   :  1464  
 3rd Qu.:   561   3rd Qu.:   853   3rd Qu.:   670   3rd Qu.:  1019  
 Max.   :188825   Max.   :196804   Max.   :172119   Max.   :328674  
## A magic trick to convert column-wise summaries into a data.frame.
## The do.call() function produces a data frame with one col per sample, 
## we transpose it to obtain one row per sample and one column per statistics.
stats.per.sample <- data.frame(t(do.call(cbind, lapply(count.table, summary))))
head(stats.per.sample)
    Min. X1st.Qu. Median Mean X3rd.Qu.   Max.
WT1    0     49.0    224  837      561 188825
WT2    0     88.0    371 1107      853 196804
WT3    0     74.2    297  903      670 172119
WT4    0    104.2    430 1464     1019 328674
WT5    0     84.0    353 1124      819 225435
WT6    0    153.0    667 2052     1599 357247
## We can now add some columns to the stats per sample
stats.per.sample$libsum <- apply(count.table, 2, sum) ## libsum
# Add some percentiles
stats.per.sample$perc05 <- apply(count.table, 2, quantile, 0.05)
stats.per.sample$perc10 <- apply(count.table, 2, quantile, 0.10)
stats.per.sample$perc90 <- apply(count.table, 2, quantile, 0.90)
stats.per.sample$perc95 <- apply(count.table, 2, quantile, 0.95)
stats.per.sample$zeros <- apply(count.table==0, 2, sum)
stats.per.sample$percent.zeros <- 100*stats.per.sample$zeros/nrow(count.table)

# View(stats.per.sample)
kable(stats.per.sample[sample(1:ncol(count.table), size = 10),],
      caption = "**Table: statistics per sample. ** We only display a random selection of 10 samples. ")
Table: statistics per sample. We only display a random selection of 10 samples.
Min. X1st.Qu. Median Mean X3rd.Qu. Max. libsum perc05 perc10 perc90 perc95 zeros percent.zeros
Snf4 0 137.0 562 1496 1268 204494 10662188 0 3.0 2802 4739 456 6.40
WT16 0 50.0 216 782 526 173478 5569418 0 1.0 1284 2388 621 8.71
WT12 0 76.0 332 1152 802 257336 8212630 0 1.0 1920 3536 560 7.86
WT23 0 79.0 345 1072 822 185052 7640884 0 2.0 1934 3416 535 7.51
Snf46 0 108.0 450 1236 1005 170179 8808358 0 2.0 2272 3984 499 7.00
Snf17 0 126.0 493 1429 1104 229969 10184815 0 2.5 2598 4460 474 6.65
Snf33 0 146.0 616 1653 1368 233323 11776986 0 3.0 3066 5280 453 6.36
Snf43 0 119.0 476 1343 1057 219417 9568585 0 2.0 2456 4182 482 6.76
Snf28 0 92.2 374 1078 850 180213 7681488 0 2.0 1957 3413 530 7.44
WT38 0 54.0 245 1010 620 270499 7197961 0 1.0 1682 3078 661 9.28

Distributions

Histograms of counts per gene

The summary only displays a few milestone values (mean, median, quartiles). In order to get a better intuition of the data, we can draw an histogram of all values.

par(mfrow=c(3,1))

hist(as.matrix(count.table), col="blue", border="white", breaks=100)

hist(as.matrix(count.table), col="blue", border="white",
     breaks=20000, xlim=c(0,2000), main="Counts per gene",
     xlab="Counts (truncated axis)", ylab="Number of genes", 
     las=1, cex.axis=0.7)

epsilon <- 1 # pseudo-count to avoid problems with log(0)
hist(as.matrix(log2(count.table + epsilon)), breaks=100, col="blue", border="white",
     main="Log2-transformed counts per gene", xlab="log2(counts+1)", ylab="Number of genes", 
     las=1, cex.axis=0.7)
Histogram of counts per genes. **Top: raw counts. ** the scale is determined by the gene with the highest count, which is apparently an outlier. **Middle: ** raw counts, with X axis truncated to 2000 in order to display a representative range despite outliers. **Bottom: ** log2-transformed counts (bottom) per gene, with a pseudocount of 1 to avoid minus infinitevalues resulting from zero counts.

Histogram of counts per genes. Top: raw counts. the scale is determined by the gene with the highest count, which is apparently an outlier. Middle: raw counts, with X axis truncated to 2000 in order to display a representative range despite outliers. Bottom: log2-transformed counts (bottom) per gene, with a pseudocount of 1 to avoid minus infinitevalues resulting from zero counts. 

par(mfrow=c(1,1))

Interpretation

  • The top histogram is not very informative so far, apparently due to the presence of a few very high count values, that impose a very large scale on the \(X\) axis.
  • The middle histogram shows the representative range. Note the height of the first bin, which includes the zero counts.
  • The logarithmic transformation (bottom histogram) improves the readability. Note that we added a pseudo-count of 1 to avoid problems with the log transformation of zero counts (which gives \(-\infty\)).

Boxplots of gene count distributions per sample

To get better insights into the distribution per sample, boxplots offer a good perspective.

## Boxplots
boxplot(log2(count.table + epsilon), col=expDesign$color, pch=".", 
        horizontal=TRUE, cex.axis=0.5,
        las=1, ylab="Samples", xlab="log2(Counts +1)")
Box plots of non-normalized log2(counts) per sample.

Box plots of non-normalized log2(counts) per sample.

Density plots

Another way to get an intuition of the value distributions is to use the geom_density() function (ggplot2 library), which draws one density curve for each sample.

## Density
## We will require some functions from the reshape2 and ggplot2 packages
if(!require("reshape2")){
  install.packages("reshape2")
}

if(!require("ggplot2")){
  install.packages("ggplot2")
}

library(ggplot2)
library(reshape2)
count_melt <- reshape2::melt(log2(count.table + epsilon))
head(count_melt)
  variable value
1      WT1  1.58
2      WT1  4.39
3      WT1  2.00
4      WT1  6.25
5      WT1  5.93
6      WT1  3.81
ggplot(data=count_melt, mapping=aes(x=value, color=variable)) + geom_density()
Densities of log2(counts). Each curve corresponds to one sample.

Densities of log2(counts). Each curve corresponds to one sample.

Beware: the R function geom_density() does not display the actual distribution of your values, but a polynomial fit. The representation thus generally looks smoother than the actual data. It is important to realize that, in some particular cases, the fit can lead to extrapolated values which can be misleading.

Scatter plots

nb.pairs <- 6


## Define a function to draw a scatter plot for a pair of variables (samples) with density colors
plotFun <- function(x,y){ 
  dns <- densCols(x,y); 
  points(x,y, col=dns, pch=".", panel.first=grid());  
#  abline(a=0, b=1, col="brown")
  }

## Plot the scatter plot for a few pairs of variables selected at random
set.seed(123) # forces the random number generator to produce fixed results. Should generally not be used, except for the sake of demonstration with a particular selection. 
pairs(log2(count.table[,sample(ncol(count.table), nb.pairs)] + epsilon), 
      panel=plotFun, lower.panel = NULL)
**Scatter plot of log2-counts for a random selection of samples. **

Scatter plot of log2-counts for a random selection of samples.

Let’s have a look at the scatter plots using the pairs() function. We will only represent 6 randomly selected samples.

The command pairs() draws a scatter plot for each pair of columns of the input dataset. The plot shows a fairly good reproducibility between samples of the same type (WT or KO, respectively): all points are aligned along te diagonal, with a relatively wider dispersion at the bottom, corresponding to small number fluctuations.

In contrast, on all the plots comparing a WT and a KO sample, we can see some points (genes) discarding from the diagonal.

Eliminating undetected genes

All genes from genome the S. cerevisiae where quantified. However only a fraction of them where expressed and some of them where to weakly expressed to be detected in any of the sample. As a result the count table may contain rows with only zero values (null counts).

  • What is the percentage of gene having null counts per sample. Draw a barplot.
  • Some genes were not detected in any of the sample. Count their number, and delete them from the count.table data.frame.

View solution| Hide solution

Selecting random samples

One of the questions that will drive the analysis will be to define the impact of the number of biological samples on the results.

The original study contained 48 replicates per genotype, what happens if we select a smaller number?

Each attendee of this course select a given number (e.g. 3, 4, 5, 10, 15, 20, 35, 40, 45…) and adapt the code below run the analysis with that number of replicates per genotype. We will at the end then compare the results (number of genes, significance, …).

nb.replicates <- 10 ## Each attendee chooses a number (3,4,5,10,15 or 20)

samples.WT <- sample(1:48, size=nb.replicates, replace=FALSE)

## Random sampling of the Snf2 replicates (columns 49 to 96)
samples.Snf2 <- sample(49:96, size=nb.replicates, replace=FALSE)

selected.samples <- c(samples.WT, samples.Snf2)

# Don't forget to update colors
col.pheno.selected <- expDesign$color[selected.samples]

Differential analysis with DESeq2

In this section we will search for genes whose expression is affected by the genetic invalidation. You will first need to install the DESeq2 bioconductor library then load it.

## Install the library if needed then load it
if (!require("BiocManager", quietly = TRUE)){
    install.packages("BiocManager")
    BiocManager::install()
}

if(!require("lazyeval")){
  install.packages("lazyeval")
}

if(!require("DESeq2")){
  BiocManager::install("DESeq2")
}

library("DESeq2")

Creating a DESeqDataSet dataset

We will then create a DESeqDataSet using the DESeqDataSetFromMatrix() function. Get some help about the DESeqDataSet and have a look at some important accessor methods: counts, conditions, estimateSizeFactors, sizeFactors, estimateDispersions and nbinomTest.

## Use the DESeqDataSetFromMatrix to create a DESeqDataSet object
dds0 <- DESeqDataSetFromMatrix(countData = count.table[,selected.samples ], colData = expDesign[selected.samples,], design = ~ strain)
print(dds0)
class: DESeqDataSet 
dim: 6887 20 
metadata(1): version
assays(1): counts
rownames(6887): 15s_rrna 21s_rrna ... ty(gua)m2 ty(gua)o
rowData names(0):
colnames(20): WT43 WT37 ... Snf36 Snf14
colData names(3): label strain color
## What kind of object is it ?
is(dds0)
[1] "DESeqDataSet"               "RangedSummarizedExperiment"
[3] "SummarizedExperiment"       "RectangularData"           
[5] "Vector"                     "Annotated"                 
[7] "vector_OR_Vector"          
isS4(dds0)
[1] TRUE
## What does it contain ?
# The list of slot names
slotNames(dds0)
[1] "design"             "dispersionFunction" "rowRanges"         
[4] "colData"            "assays"             "NAMES"             
[7] "elementMetadata"    "metadata"          
## Get some help about the "CountDataSet" class.
## NOT RUN
#?"DESeqDataSet-class"

Normalization

The normalization procedure (RLE) is implemented through the estimateSizeFactors function.

How is the scaling factor computed ?

Given a matrix with \(p\) columns (samples) and \(n\) rows (genes) this function estimates the size factors as follows: Each column element is divided by the geometric means of the rows. For each sample, the median (or, if requested, another location estimator) of these ratios (skipping the genes with a geometric mean of zero) is used as the size factor for this column.

The scaling factor for sample \(j\) is thus obtained as:

\[sf_{j} = median(\frac{K_{g,j}}{(\prod_{j=1}^p K_{g,j})^{1/p}}) \]

### Let's implement such a function
### cds is a countDataset
estimSf <- function (cds){
    # Get the count matrix
    cts <- counts(cds)
    
    # Compute the geometric mean
    geomMean <- function(x) prod(x)^(1/length(x))

    # Compute the geometric mean over the line
    gm.mean  <-  apply(cts, 1, geomMean)
    
    # Zero values are set to NA (avoid subsequentcdsdivision by 0)
    gm.mean[gm.mean == 0] <- NA
    
    # Divide each line by its corresponding geometric mean
    # sweep(x, MARGIN, STATS, FUN = "-", check.margin = TRUE, ...)
    # MARGIN: 1 or 2 (line or columns)
    # STATS: a vector of length nrow(x) or ncol(x), depending on MARGIN
    # FUN: the function to be applied
    cts <- sweep(cts, 1, gm.mean, FUN="/")
    
    # Compute the median over the columns
    med <- apply(cts, 2, median, na.rm=TRUE)
    
    # Return the scaling factor
    return(med)
}

Now, check that the results obtained with our function are the same as those produced by DESeq. The method associated with normalization for the “CountDataSet” class is estimateSizeFactors().

## Normalizing using the method for an object of class"CountDataSet" 
dds.norm <-  estimateSizeFactors(dds0)
sizeFactors(dds.norm)
 WT43  WT37  WT14  WT25  WT26  WT27   WT5  WT48  WT28   WT9 Snf29 Snf35  Snf8 
0.720 0.889 1.070 1.161 0.945 1.018 0.856 0.828 1.139 0.862 1.044 0.923 0.941 
Snf26  Snf7 Snf42  Snf9 Snf19 Snf36 Snf14 
1.274 1.664 0.825 1.234 0.890 0.946 1.242 
## Now get the scaling factor with our homemade function.cds.norm
head(estimSf(dds0)) 
 WT43  WT37  WT14  WT25  WT26  WT27 
0.720 0.889 1.070 1.161 0.945 1.018 
all(round(estimSf(dds0),6) == round(sizeFactors(dds.norm), 6))
[1] TRUE
## Checking the normalization
par(mfrow=c(1,2),cex.lab=0.7)

boxplot(log2(counts(dds.norm)+epsilon),  col=col.pheno.selected, cex.axis=0.7, 
        las=1, xlab="log2(counts)", horizontal=TRUE, main="Raw counts")
boxplot(log2(counts(dds.norm, normalized=TRUE)+epsilon),  col=col.pheno.selected, cex.axis=0.7, 
        las=1, xlab="log2(normalized counts)", horizontal=TRUE, main="Normalized counts")
**Impact of the count normalization. **

Impact of the count normalization. 

if(!require("patchwork")){
  install.packages("patchwork")
}

p1 <- ggplot(data=count_melt, mapping=aes(x=value, color=variable)) + geom_density() +  theme(legend.position = "none")
count_norm_melt <- melt(log2(counts(dds.norm, normalized=TRUE)+epsilon))
head(count_norm_melt)
      Var1 Var2 value
1 15s_rrna WT43  4.03
2 21s_rrna WT43  6.66
3     hra1 WT43  1.92
4     icr1 WT43  7.30
5     lsr1 WT43  7.32
6     nme1 WT43  3.90
p2 <- ggplot(data=count_norm_melt, mapping=aes(x=value, color=Var2)) + geom_density() + theme(legend.position = "none")
p1 + p2

Modeling read counts

Let us imagine that we would produce a lot of RNA-Seq experiments from the same samples (technical replicates). For each gene \(g\) the measured read counts would be expected to vary rather slighlty around the expected mean and would be probably well modeled using a Poisson distribution. However, when working with biological replicates more variations are intrinsically expected. Indeed, the measured expression values for each genes are expected to fluctuate more importantly, due to the combination of biological and technical factors: inter-individual variations in gene regulation, sample purity, cell-synchronization issues or reponses to environment (e.g. heat-shock).

The Poisson distribution has only one parameter indicating its expected mean : \(\lambda\). The variance of the distribution equals its mean \(\lambda\). Thus in most cases, the Poisson distribution is not expected to fit very well with the count distribution in biological replicates, since we expect some over-dispersion (greater variability) due to biological noise.

As a consequence, when working with RNA-Seq data, many of the current approaches for differential expression call rely on an alternative distribution: the negative binomial (note that this holds true also for other -Seq approaches, e.g. ChIP-Seq with replicates).

What is the negative binomial ?

The negative binomial distribution is a discrete distribution that give us the probability of observing \(x\) failures before a target number of succes \(n\) is obtained. As we will see later the negative binomial can also be used to model over-dispersed data (in this case this overdispersion is relative to the poisson model).

The probability of \(x\) failures before \(n\) success

First, given a Bernouilli trial with a probability \(p\) of success, the negative binomial distribution describes the probability of observing \(x\) failures before a target number of successes \(n\) is reached. In this case the parameters of the distribution will thus be \(p\), \(n\) (in dnbinom() function of R, \(n\) and \(p\) are denoted by arguments size and prob respectively).

\[P_{NegBin}(x; n, p) = \binom{x+n-1}{x}\cdot p^n \cdot (1-p)^x = C^{x}_{x+n-1}\cdot p^n \cdot (1-p)^x \]

In this formula, \(p^n\) denotes the probability to observe \(n\) successes, \((1-p)^x\) the probability of \(x\) failures, and the binomial coefficient \(C^{x}_{x+n-1}\) indicates the number of possible ways to dispose \(x\) failures among the \(x+n-1\) trials that precede the last one (the problem statement imposes for the last trial to be a success).

The negative binomial distribution has expected value \(n\frac{q}{p}\) and variance \(n\frac{q}{p^2}\). Some examples of using this distribution in R are provided below.

Particular case: when \(n=1\) the negative binomial corresponds to the the geometric distribution, which models the probability distribution to observe the first success after \(x\) failures: \(P_{NegBin}(x; 1, p) = P_{geom}(x; p) = p \cdot (1-p)^x\).

par(mfrow=c(1,1))

## Some intuition about the negative binomiale parametrized using n and p.
## The simple case, one success (see geometric distribution)
# Let's have a look at the density
p <- 1/6 # the probability of success
n <- 1   # target for number of successful trials

# The density function
plot(0:10, dnbinom(0:10, n, p), type="h", col="blue", lwd=4)
Negative binomial distribution.

Negative binomial distribution. 

# the probability of zero failure before one success.
# i.e the probability of success 
dnbinom(0, n , p)
[1] 0.167
## i.e the probability of at most 5 failure before one success. 
sum(dnbinom(0:5, n , p)) # == pnbinom(5, 1, p)
[1] 0.665
## The probability of at most 10 failures before one sucess 
sum(dnbinom(0:10, n , p)) # == pnbinom(10, 1, p)
[1] 0.865
## The probability to have more than 10 failures before one sucess
1-sum(dnbinom(0:10, n , p)) # == 1 - pnbinom(10, 1, p)
[1] 0.135
## With two successes
## The probability of x failure before two success (e.g. two six)
n <- 2
plot(0:30, dnbinom(0:30, n, p), type="h", col="blue", lwd=2,
     main="Negative binomial density",
     ylab="P(x; n,p)",
     xlab=paste("x = number of failures before", n, "successes"))

# Expected value
q <- 1-p
(ev <- n*q/p)
[1] 10
abline(v=ev, col="darkgreen", lwd=2)

# Variance 
(v <- n*q/p^2)
[1] 60
arrows(x0=ev-sqrt(v), y0 = 0.04, x1=ev+sqrt(v), y1=0.04, col="brown",lwd=2, code=3, , length=0.2, angle=20)
Negative binomial distribution.

Negative binomial distribution.

Using mean and dispersion

The second way of parametrizing the distribution is using the mean value \(m\) and the dispersion parameter \(r\) (in dnbinom() function of R, \(m\) and \(r\) are denoted by arguments mu and size respectively). The variance of the distribution can then be computed as \(m + m^2/r\).

Note that \(m\) can be deduced from \(n\) and \(p\).

n <- 10
p <- 1/6
q <- 1-p
mu <- n*q/p

all(dnbinom(0:100, mu=mu, size=n) == dnbinom(0:100, size=n, prob=p))
[1] FALSE

Modelling read counts through a negative binomial

To perform diffential expression call DESeq will assume that, for each gene, the read counts are generated by a negative binomial distribution. One problem here will be to estimate, for each gene, the two parameters of the negative binomial distribution: mean and dispersion.

  • The mean will be estimated from the observed normalized counts in both conditions.

  • The first step will be to compute a gene-wise dispersion. When the number of available samples is insufficient to obtain a reliable estimator of the variance for each gene, DESeq will apply a shrinkage strategy, which assumes that counts produced by genes with similar expression level (counts) have similar variance (note that this is a strong assumption). DESeq will regress the gene-wise dispersion onto the means of the normalized counts to obtain an estimate of the dispersion that will be subsequently used to build the binomial model for each gene.

## Performing estimation of dispersion parameter
dds.disp <- estimateDispersions(dds.norm)

## A diagnostic plot which
## shows the mean of normalized counts (x axis)
## and dispersion estimate for each genes
plotDispEsts(dds.disp)

Performing differential expression call

Now that a negative binomial model has been fitted for each gene, the nbinomWaldTest can be used to test for differential expression. The output is a data.frame which contains nominal p-values, as well as FDR values (correction for multiple tests computed with the Benjamini–Hochberg procedure).

alpha <- 0.0001
wald.test <- nbinomWaldTest(dds.disp)
res.DESeq2 <- results(wald.test, alpha=alpha, pAdjustMethod="BH")

## What is the object returned by nbinomTest()
class(res.DESeq2)
[1] "DESeqResults"
attr(,"package")
[1] "DESeq2"
is(res.DESeq2) # a data.frame
 [1] "DESeqResults"      "DFrame"            "DataFrame"        
 [4] "SimpleList"        "RectangularData"   "List"             
 [7] "DataFrame_OR_NULL" "Vector"            "list_OR_List"     
[10] "Annotated"         "vector_OR_Vector" 
slotNames(res.DESeq2)
[1] "priorInfo"       "rownames"        "nrows"           "elementType"    
[5] "elementMetadata" "metadata"        "listData"       
head(res.DESeq2)
log2 fold change (MLE): strain WT vs Snf 
Wald test p-value: strain WT vs Snf 
DataFrame with 6 rows and 6 columns
          baseMean log2FoldChange     lfcSE      stat      pvalue        padj
         <numeric>      <numeric> <numeric> <numeric>   <numeric>   <numeric>
15s_rrna  28.08058      -1.240636 0.8926980 -1.389760          NA          NA
21s_rrna 128.41370      -0.572078 0.7000361 -0.817213          NA          NA
hra1       2.29627       0.443593 0.5452527  0.813555 0.415900041 0.529763145
icr1     147.68494      -0.346711 0.0913234 -3.796520 0.000146742 0.000559623
lsr1     224.52147      -0.235851 0.4436689 -0.531591 0.595008983 0.696343313
nme1      29.00707      -0.567462 0.3515400 -1.614219 0.106480031 0.179555630
## The column names of the data.frame
## Note the column padj 
## contains FDR values (computed Benjamini–Hochberg procedure)
colnames(res.DESeq2)
[1] "baseMean"       "log2FoldChange" "lfcSE"          "stat"          
[5] "pvalue"         "padj"          
## Order the table by decreasing p-valuer
res.DESeq2 <- res.DESeq2[order(res.DESeq2$padj),]
head(res.DESeq2)
log2 fold change (MLE): strain WT vs Snf 
Wald test p-value: strain WT vs Snf 
DataFrame with 6 rows and 6 columns
         baseMean log2FoldChange     lfcSE      stat       pvalue         padj
        <numeric>      <numeric> <numeric> <numeric>    <numeric>    <numeric>
ygr234w  2965.003        4.12298 0.0939529   43.8835  0.00000e+00  0.00000e+00
yhr215w  1298.565        4.37142 0.1163141   37.5829  0.00000e+00  0.00000e+00
yil121w   926.134        2.74561 0.0662922   41.4169  0.00000e+00  0.00000e+00
yor290c   853.322        7.41602 0.1787839   41.4803  0.00000e+00  0.00000e+00
yar071w  1666.340        4.13109 0.1111865   37.1546 3.69040e-302 4.86395e-299
ydr033w  4051.107        3.60699 0.1045447   34.5019 7.51713e-261 8.25631e-258
## Draw an histogram of the p-values
hist(res.DESeq2$padj, breaks=20, col="grey", main="DESeq2 p-value distribution", xlab="DESeq2 P-value", ylab="Number of genes")
Histogram of the p-values reported by DESeq2.

Histogram of the p-values reported by DESeq2.

Volcano plot

alpha <- 0.01 # Threshold on the adjusted p-value
cols <- densCols(res.DESeq2$log2FoldChange, -log10(res.DESeq2$pvalue))
plot(res.DESeq2$log2FoldChange, -log10(res.DESeq2$padj), col=cols, panel.first=grid(),
     main="Volcano plot", xlab="Effect size: log2(fold-change)", ylab="-log10(adjusted p-value)",
     pch=20, cex=0.6)
abline(v=0)
abline(v=c(-1,1), col="brown")
abline(h=-log10(alpha), col="brown")

gn.selected <- abs(res.DESeq2$log2FoldChange) > 2 & res.DESeq2$padj < alpha 
text(res.DESeq2$log2FoldChange[gn.selected],
     -log10(res.DESeq2$padj)[gn.selected],
     lab=rownames(res.DESeq2)[gn.selected ], cex=0.4)
Volcano plot of DESeq2 results. Abcsissa: log2(fold-change). Ordinate: significance ($-log_{10}(P-value)$).

Volcano plot of DESeq2 results. Abcsissa: log2(fold-change). Ordinate: significance (\(-log_{10}(P-value)\)).

Check the expression levels of the most differentially expressed gene

It may be important to check the validity of our analysis by simply assessing the expression level of the most highly differential gene.

gn.most.sign <- rownames(res.DESeq2)[1]
gn.most.diff.val <- counts(dds.norm, normalized=T)[gn.most.sign,]
barplot(gn.most.diff.val, col=col.pheno.selected, main=gn.most.sign, las=2, cex.names=0.5)
Barplot of the counts per sample fr a selected gene.

Barplot of the counts per sample fr a selected gene.

Looking at the results with a MA plot

One popular diagram in dna chip analysis is the M versus A plot (MA plot) between two conditions \(a\) and \(b\). In this representation :

  • M (Minus) is the log ratio of counts calculated for any gene. \[M_g = log2(\bar{x}_{g,a}) - log2(\bar{x}_{g,b})\]
  • A (add) is the average log counts which corresponds to an estimate of the gene expression level. \[A_g = \frac{1}{2}(log2(\bar{x}_g,a) + log2(\bar{x}_g,b))\]
## Draw a MA plot.
## Genes with adjusted p-values below 1% are shown
plotMA(res.DESeq2, colNonSig = "blue")
abline(h=c(-1:1), col="red")
MA plot. The abcsissa indicates the mean of normalized counts; the ordinate the log2(fold-change).

MA plot. The abcsissa indicates the mean of normalized counts; the ordinate the log2(fold-change).

Hierarchical clustering

To ensure that the selected genes distinguish well between “treated”” and “untreated” condition we will perform a hierachical clustering using the heatmap.2() function from the gplots library.

## We select gene names based on FDR (1%)
gene.kept <- rownames(res.DESeq2)[res.DESeq2$padj <= alpha & !is.na(res.DESeq2$padj)]

## We retrieve the normalized counts for gene of interest
count.table.kept <- log2(count.table + epsilon)[gene.kept, ]
dim(count.table.kept)
[1] 2398   96
## Install the gplots library if needed then load it
if(!require("gplots")){
  install.packages("gplots")
}
library("gplots")

## Perform the hierarchical clustering with
## A distance based on Pearson-correlation coefficient
## and average linkage clustering as agglomeration criteria
heatmap.2(as.matrix(count.table.kept), 
          scale="row", 
          hclust=function(x) hclust(x,method="average"), 
          distfun=function(x) as.dist((1-cor(t(x)))/2), 
          trace="none", 
          density="none", 
          labRow="",
          cexCol=0.7)
Heatmap of the gebes deckared significant with DESeq2. Rows correspond to genes, columns to samples.

Heatmap of the gebes deckared significant with DESeq2. Rows correspond to genes, columns to samples.

Functional enrichment

We will now perform functional enrichment using the list of induced genes. This step will be performed using the gProfileR R library.

library(gProfileR)

res.DESeq2.df <- na.omit(data.frame(res.DESeq2))
induced.sign <- rownames(res.DESeq2.df)[res.DESeq2.df$log2FoldChange >= 2 &  res.DESeq2.df$padj < alpha]
# head(induced.sign)
# names(term.induced)

term.induced <- gprofiler(query=induced.sign, organism="scerevisiae")
term.induced <- term.induced[order(term.induced$p.value),]
# term.induced$p.value
kable(term.induced[1:10,c("term.name",
                      "term.size",
                      "query.size",
                      "overlap.size",
                      "recall",
                      "precision",
                      "p.value", 
                      "intersection")], 
      format.args=c(engeneer=TRUE, digits=3), caption="**Table: functional analysis wit gProfileR. ** ")
Table: functional analysis wit gProfileR.
term.name term.size query.size overlap.size recall precision p.value intersection
39 RNA-DNA hybrid ribonuclease activity 48 85 30 0.625 0.353 0 YAR009C,YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B
10 DNA integration 50 85 30 0.600 0.353 0 YAR009C,YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B
49 RNA-directed DNA polymerase activity 52 85 30 0.577 0.353 0 YAR009C,YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B
42 aspartic-type peptidase activity 54 85 29 0.537 0.341 0 YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B
44 aspartic-type endopeptidase activity 54 85 29 0.537 0.341 0 YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B
38 endoribonuclease activity, producing 5’-phosphomonoesters 63 85 30 0.476 0.353 0 YAR009C,YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B
50 DNA-directed DNA polymerase activity 63 85 30 0.476 0.353 0 YAR009C,YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B
48 DNA polymerase activity 67 85 30 0.448 0.353 0 YAR009C,YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B
35 endonuclease activity, active with either ribo- or deoxyribonucleic acids and producing 5’-phosphomonoesters 74 85 30 0.405 0.353 0 YAR009C,YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B
37 endoribonuclease activity 77 85 30 0.390 0.353 0 YAR009C,YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B

And now using the list of repressed genes.

res.DESeq2.df <- na.omit(data.frame(res.DESeq2))
repressed.sign <- rownames(res.DESeq2.df)[res.DESeq2.df$log2FoldChange <= -2 &  res.DESeq2.df$padj < alpha]
head(repressed.sign)
[1] "yer081w" "ypl025c" "ygr051c" "ycr025c" "ypr064w" "ygr087c"
term.repressed <- gprofiler(query=repressed.sign, organism="scerevisiae")
term.repressed <- term.repressed[order(term.repressed$p.value),]
kable(head(term.induced[,c("p.value", "term.name","intersection")], 10))
p.value term.name intersection
39 0 RNA-DNA hybrid ribonuclease activity YAR009C,YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B
10 0 DNA integration YAR009C,YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B
49 0 RNA-directed DNA polymerase activity YAR009C,YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B
42 0 aspartic-type peptidase activity YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B
44 0 aspartic-type endopeptidase activity YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B
38 0 endoribonuclease activity, producing 5’-phosphomonoesters YAR009C,YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B
50 0 DNA-directed DNA polymerase activity YAR009C,YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B
48 0 DNA polymerase activity YAR009C,YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B
35 0 endonuclease activity, active with either ribo- or deoxyribonucleic acids and producing 5’-phosphomonoesters YAR009C,YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B
37 0 endoribonuclease activity YAR009C,YBL005W-B,YBR012W-B,YDR098C-B,YDR210C-D,YDR261C-D,YDR316W-B,YDR365W-B,YER138C,YER160C,YGR027W-B,YGR038C-B,YGR161C-D,YHR214C-B,YJR027W,YJR029W,YLR035C-A,YLR157C-B,YLR227W-B,YML039W,YML045W,YMR045C,YMR050C,YNL284C-B,YOL103W-B,YOR142W-B,YPL257W-B,YPR137C-B,YPR158C-D,YPR158W-B

Assess the effect of sample number on differential expression call

Using a loop, randomly select 10 times 2,5,10,15..45 samples from WT and Snf2 KO. Perform differential expression calls and draw a diagram showing the number of differential expressed genes.

## Create a directory to store the results that will be obtained below
dir.results <- file.path(dir.snf2, "results")
dir.create(dir.results, showWarnings = FALSE, recursive = TRUE)

## Export the table with statistics per sample.
write.table(stats.per.sample, file=file.path(dir.results, "stats_per_sample.tsv"),
            quote=FALSE, sep="\t", col.names =NA, row.names = TRUE)

# Export the DESeq2 result table
DESeq2.table <- file.path(dir.results, "yeast_Snf2_vs_WT_DESeq2_diff.tsv")
write.table(res.DESeq2, file=DESeq2.table, col.names = NA, row.names = TRUE, sep="\t", quote = FALSE)

  1. A factor is a vector with levels (categories), which permits an efficient storage and indexing, but can in some cases lead to misleading effects. To circumvent this, we will sometimes have to convert the factor to a vector, with the R command as.vector(). ↩︎