Simple workflow for isobaric-labeling proteomic data with ANOVA, t-testing, DEqMS/limma and annotation via fgsea
The author recommends RStudio IDE. The script was developed and tested on Ubuntu 20.04 with R 3.6.
Usage:
- Download and run the ".R" script, adapting the paths and formats
- Pay particular attention to the format of the output file and to sample names. The example table has the quantitative data in "Abundance Ratio" columns, the biological replicates from the same group are named xyz_1, xyz_2, xyz_3.
See the main highlights of the workflows below:
#library(readxl)
library(dplyr)
library(ggplot2)
library(tidyr)
library(DEqMS)
library(fgsea)Select the work directory and import the protein table.
The example data originates from the analysis of the commercial yeast triple-knockout (TKO) TMT standard
LC-MS rw files were processed via Proteome Discoverer 2.4, the output files were saved as tab-separated text file with "R-friendly header" set to "True".
#Set the working directory for the project
setwd("/home/trainee/Documents/R/Yeast_TKO/")
#If the table has been exported as a tab-separated text file
all_proteins <- read.table("MS3_FAIMS_Proteins_PD24.txt", sep = '\t', header = TRUE)
#If the table is in Excel format
#all_proteins <- read_excel("MS3_FAIMS_Proteins_PD24.xlsx")dim(all_proteins)## [1] 1904 51
The quantitative columns (abundance ratio columns) are extracted and renamed:
res <- rename_ratios(all_proteins)
all_proteins <- res[[1]]
quan_columns <- res[[2]]
quan_columns## [1] "BY4741_2" "his4_1" "his4_2" "his4_3" "met6_1" "met6_2" "met6_3" "ura2_1" "ura2_2" "ura2_3"
After several stages of filtering, we check the abundance distibution in each sample:
dfWide <- all_proteins %>%
filter(!grepl("cont_",Accession)) %>%
subset (select=c("Accession", quan_columns) ) %>%
na.omit()
rownames(dfWide) <- dfWide$Accession
dfWide$Accession <- NULL
dfWide <- log2(dfWide)
#Look at the distribution of quan values
summary(dfWide)## BY4741_2 his4_1 his4_2 his4_3 met6_1
## Min. :-0.76121 Min. :-2.426625 Min. :-2.343732 Min. :-2.434403 Min. :-3.36587
## 1st Qu.:-0.08161 1st Qu.:-0.189351 1st Qu.:-0.193058 1st Qu.:-0.198636 1st Qu.:-0.23447
## Median :-0.01013 Median :-0.020340 Median :-0.014500 Median :-0.017417 Median :-0.08238
## Mean :-0.01831 Mean : 0.005853 Mean : 0.003676 Mean : 0.000839 Mean :-0.06536
## 3rd Qu.: 0.04963 3rd Qu.: 0.180148 3rd Qu.: 0.186501 3rd Qu.: 0.180148 3rd Qu.: 0.09390
## Max. : 0.76043 Max. : 3.528321 Max. : 3.441218 Max. : 3.478195 Max. : 1.84679
## met6_2 met6_3 ura2_1 ura2_2 ura2_3
## Min. :-3.15843 Min. :-3.26534 Min. :-2.49005 Min. :-2.57347 Min. :-2.44222
## 1st Qu.:-0.20257 1st Qu.:-0.22600 1st Qu.:-0.18033 1st Qu.:-0.18115 1st Qu.:-0.16974
## Median :-0.06492 Median :-0.07704 Median :-0.03579 Median :-0.04097 Median :-0.03357
## Mean :-0.04076 Mean :-0.05791 Mean :-0.02990 Mean :-0.03511 Mean :-0.02894
## 3rd Qu.: 0.10467 3rd Qu.: 0.09119 3rd Qu.: 0.12433 3rd Qu.: 0.11636 3rd Qu.: 0.10836
## Max. : 1.68840 Max. : 1.79119 Max. : 2.07108 Max. : 1.61259 Max. : 1.92828
#Box Plot
boxplot(
Log2_Abund~Sample, data = gather(dfWide, Sample, Log2_Abund),
main = "Original Log2 Ratios"
)
Proteomic data may require normalization. Let's normalize each sample on it's median and check out the resulting distributions:
#For each column, subtract the median of the column from each of it's values
dfNorm <- mapply('-', dfWide, apply(dfWide,2,median))
#Transform into a dataframe
dfNorm <- as.data.frame(dfNorm, row.names = row.names(dfWide))
boxplot(
Log2_Abund~Sample, data = gather(dfNorm, Sample, Log2_Abund),
main = "Normalized Log2 Ratios"
)
#If desired, use the normalized data for the downstream operations
dfWide <- dfNorm
dfWide$Accession <- row.names(dfWide)
dfLong <- gather(dfWide, Sample, Log2_Abund, -Accession)
dfWide$Accession <- NULLLook at the principle component analysis on samples:
dfWide.t <- t(dfWide)
dfWide.pca <- prcomp(dfWide.t, center = TRUE, scale. = FALSE)
summary(dfWide.pca)## Importance of components:
## PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8 PC9 PC10
## Standard deviation 4.7762 3.8970 2.8019 1.78159 1.61259 1.53323 1.51155 1.40213 1.36160 2.443e-15
## Proportion of Variance 0.3797 0.2528 0.1307 0.05283 0.04328 0.03913 0.03803 0.03272 0.03086 0.000e 00
## Cumulative Proportion 0.3797 0.6325 0.7631 0.81598 0.85926 0.89839 0.93642 0.96914 1.00000 1.000e 00
Let's plot the first 4 principal components, as they explain the bulk of the variation within the data set:
dfWide.pca <- as.data.frame(dfWide.pca$x)
dfWide.pca$Group <- sapply(
as.character( row.names(dfWide.pca) ),
function(x) {
strsplit(x, "_")[[1]][[1]]
}
)
#Principal components 1 and 2
ggplot(
dfWide.pca,
aes(x = PC1, y = PC2, colour = Group )
)
geom_point(shape=19, size=4, alpha = 0.7)
geom_hline(yintercept = 0, colour = "gray65")
geom_vline(xintercept = 0, colour = "gray65")
ggtitle("PCA On Proteins")
theme_classic()
#Principal components 3 and 4
ggplot(
dfWide.pca,
aes(x = PC3, y = PC4, colour = Group )
)
geom_point(shape=19, size=4, alpha = 0.7)
geom_hline(yintercept = 0, colour = "gray65")
geom_vline(xintercept = 0, colour = "gray65")
ggtitle("PCA On Proteins")
theme_classic()
One-way ANOVA analysis:
cols_anova <- c("his4_1", "his4_2", "his4_3", "met6_1", "met6_2", "met6_3", "ura2_1", "ura2_2", "ura2_3")
#Create the named list with our groups
groups <- list()
for (i in cols_anova) {
local_name <- strsplit(i, "_")[[1]][[1]]
if ( local_name %in% names(groups) ) {
groups[[local_name]] <- c(groups[[local_name]], i)
} else {
groups[[local_name]] <- c(i)
}
}
groups## $his4
## [1] "his4_1" "his4_2" "his4_3"
##
## $met6
## [1] "met6_1" "met6_2" "met6_3"
##
## $ura2
## [1] "ura2_1" "ura2_2" "ura2_3"
dfANOVA <- dfWide
dfANOVA$anovaPval <- apply(dfANOVA, 1, function(x) {
df <- as.data.frame(x)
#Select the samples for ANOVA
#Important if you need to exclude some of the samples from the calculation
#cols_anova <- c("his4_1", "his4_2", "his4_3", "met6_1", "met6_2", "met6_3", "ura2_1", "ura2_2", "ura2_3")
df$Sample <- rownames(df)
df <- df[ df$Sample %in% cols_anova, ]
#Define groups in sync with the selected columns
#OBS: format-dependent
df$Group <- as.vector(
sapply(
cols_anova,
function(x) { strsplit(x, "_")[[1]][[1]] }
)
)
anovaResults <- aov(x ~ Group, data = df)
#This Very exciting expression is how to extract the p-value from the aov summary
return(summary(anovaResults)[[1]]["Pr(>F)"][[1]][[1]])
})
#Benjamini-Hochberg correction for multiple testing
dfANOVA$adjPval <- p.adjust(dfANOVA$anovaPval, method = "BH")
#Add group averages
for ( i in names(groups) ) {
dfANOVA[i] <- apply(
dfANOVA, 1, function(x) {
#print(x)
#print(typeof(x))
mean( x[ groups[[i]] ] )
}
)
}
#Add max LogFC
dfANOVA$MaxLog2FC <- apply(
dfANOVA, 1, function(x) {
log_means <- x[ names(groups) ]
max(log_means) - min(log_means)
}
)Select proteins with adjusted ANOVA P <= 0.05 and FC > log2(1.3) (30%)
dfANOVA.Sign <- dfANOVA %>%
filter(adjPval <= 0.05 & MaxLog2FC >= log2(1.3) ) %>%
select(cols_anova)
dim(dfANOVA.Sign)## [1] 190 9
heatmap(
as.matrix( dfANOVA.Sign ), scale = "none",
col = colorRampPalette(c("red", "grey", "blue"))(256)
)
The classic way to compare levels in two groups is the Student's t-test.
Let's compare the met6 knockouts and his4 knockouts as an example:
calc_ttest <- function(df, groupping, gr1, gr2, maxAdjP, minFC) {
df <- df[ c( groupping[[gr1]], groupping[[gr2]] ) ]
#Log2 fold change group2 - group1
df$Log2FC <- apply(
df, 1, function(x) {
mean( x[ groupping[[gr2]] ] ) - mean( x[ groupping[[gr1]] ] )
}
)
#T-test with equal variance
df$T_Pval <- apply(
df, 1, function(x) {
res <- t.test(
x[ groupping[[gr2]] ], x[ groupping[[gr1]] ],
alternative = "two.sided", var.equal = TRUE
)
mean( x[ groupping[[gr2]] ] ) - mean( x[ groupping[[gr1]] ] )
res$p.value
}
)
#Benjamini-Hochberg correction for multiple testing
df$adjPval <- p.adjust(df$T_Pval, method = "BH")
df$Log10adjPval <- -1*log10(df$adjPval)
#Add the categorical column
df$Diff_Abund <- apply(
df, 1, function(x) {
if (x[["adjPval"]] <= maxAdjP & x[["Log2FC"]] >= minFC) {
return( paste("Up in", gr2) )
} else if (x[["adjPval"]] <= maxAdjP & x[["Log2FC"]] <= -1*minFC) {
return( paste("Up in", gr1) )
} else {
return('Non-significant')
}
}
)
df
}
maxAdjP <- 0.05
minLog2FC <- round(log2(1.3), 3)
gr1 <- "met6"
gr2 <- "his4"
dfTtest <- calc_ttest(dfWide, groups, gr1, gr2, maxAdjP, minLog2FC )
#How many proteins have adj. P <= 0.05?
dim( filter(dfTtest, adjPval <= maxAdjP) )## [1] 493 11
Let's ad the minimal fold change to concentrate on the strongest significant effects.
There were 116 proteins that pass the filtering criteria:
dim(dfTtest %>%
filter(adjPval <= maxAdjP) %>%
filter( Log2FC >= minLog2FC | Log2FC <= -1*minLog2FC ))## [1] 116 11
Volcano plot for the met6-his4 comparison:
#Add columns with the gene names and other info
dfTtest <- merge(
dfTtest, all_proteins,
by.x="row.names", by.y="Accession",
suffixes=c("", "_"), sort=FALSE
)
#Volcano plot
ggplot(
dfTtest,
aes(x = Log2FC, y = Log10adjPval, colour = Diff_Abund )
)
geom_point(shape=19, size=2, alpha = 0.6)
geom_hline(yintercept = -1*log10(maxAdjP), colour = "gray65")
geom_vline(xintercept = 0, colour = "gray65")
geom_vline(xintercept = -1*minLog2FC, colour = "gray65")
geom_vline(xintercept = minLog2FC, colour = "gray65")
ggtitle(
paste(
"T-test ", gr1, " vs ", gr2,
" Adjusted P-value<=", maxAdjP, " Log2 FC>=", minLog2FC,
sep=""
)
)
theme_classic()
theme(
legend.title = element_blank(), legend.text = element_text(size=12),
plot.title = element_text(size=16)
)
labs(x = paste("Log2 FC", gr2, "-", gr1), y = "-Log10 Adj. P-value" )
geom_text(
data = subset(dfTtest, Log2FC >=0.9 | Log2FC <= -0.8),
aes( Log2FC, Log10adjPval, label = Gene),
alpha = 0.6, hjust = 0.5, vjust = -0.6
)
DEqMS is an R package for differential expression analysis that have been developped specifically for proteomic data by Yafeng Zhu in Janne Lehtiö's lab in Stockholm. The algorithm estimates the protein variance based on the number of peptides or PSMs, claiming better accuracy in detecting differences in protein levels, see the open-access MCP article for details. DEqMS builds on the widely used limma package, p-value and BH-adjusted p-value from limma are reported as a bonus, alongside the spectra count adjusted posterior p-values and BH-adjusted p-values calculated by DEqMS.
Let's apply the algorithm to the TKO data:
dfD <- dfWide[cols_anova]
#Define the design vector
cond = as.factor(
c("his4", "his4", "his4", "met6", "met6", "met6", "ura2", "ura2", "ura2")
)
design = model.matrix(~0 cond)
colnames(design) = gsub("cond","",colnames(design))
#Make contrasts
x <- c(
"his4-met6", "his4-ura2", "ura2-met6"
)
contrast = makeContrasts(contrasts=x,levels=design)
fit1 <- lmFit(dfD, design)
fit2 <- contrasts.fit(fit1,contrasts = contrast)
fit3 <- eBayes(fit2)
#Extract PSM count information
psm_count_table <- dfD %>%
merge(
all_proteins[c("Accession", "Number.of.PSMs")],
by.x="row.names", by.y="Accession", suffixes=c("", "_"), sort=FALSE
)
row.names(psm_count_table) <- psm_count_table$Row.names
psm_count_table <- psm_count_table[c("Number.of.PSMs")]
fit3$count = psm_count_table[rownames(fit3$coefficients),"Number.of.PSMs"]
fit4 = spectraCounteBayes(fit3)The data-dependent variance distribution looks as follows:
VarianceBoxplot(
fit4, n=30, main="TKO Variance according to DEqMS", xlab="PSM count"
)
Let's look at the contrast ura2-met6, column number 3:
current_contrast <- colnames(fit4$coefficients)[[1]]
res_DEqMS = outputResult(fit4, coef_col = 1)
head(res_DEqMS)## logFC AveExpr t P.Value adj.P.Val B gene count sca.t
## P00815 -2.8532509 -0.54962113 -85.67738 8.428621e-13 1.318236e-09 18.82971 P00815 20 -108.89141
## P05694 3.5999485 -0.85859918 77.41234 1.846004e-12 1.443575e-09 18.37417 P05694 30 96.37396
## P37291 -0.6922910 0.40048261 -28.07419 4.585032e-09 1.434198e-06 11.77468 P37291 36 -42.07175
## P15992 -0.6353142 0.44930312 -27.22447 5.803328e-09 1.512734e-06 11.53513 P15992 18 -36.62270
## P39954 -0.5661252 0.01364542 -22.95737 2.138660e-08 3.344865e-06 10.18583 P39954 22 -31.96714
## Q12443 -1.2662064 0.60896317 -44.48587 1.329146e-10 6.929282e-08 15.15264 Q12443 4 -30.06690
## sca.P.Value sca.adj.pval
## P00815 8.998668e-18 1.407392e-14
## P05694 3.360470e-17 2.627888e-14
## P37291 2.531791e-13 1.319907e-10
## P15992 1.121588e-12 4.385408e-10
## P39954 4.810014e-12 1.504572e-09
## Q12443 9.260675e-12 2.413949e-09
Add negative log10-transfromed adjusted p-values for plotting:
res_DEqMS$log.adj.P.Val = -log10(res_DEqMS$adj.P.Val )
res_DEqMS$log.sca.adj.pval = -log10(res_DEqMS$sca.adj.pval)How many proteins have adj. P <= 0.05 according to LIMMA?
maxAdjP <- 0.05
minLog2FC <- round(log2(1.3), 3)
dim(
filter(res_DEqMS, adj.P.Val <= maxAdjP)
)## [1] 611 13
How many proteins have adj. P <= 0.05 and Log2 FC >= log2(1.3) [30%] according to LIMMA?
dim(res_DEqMS %>%
filter(adj.P.Val <= maxAdjP) %>%
filter( logFC >= minLog2FC | logFC <= -1*minLog2FC ))## [1] 148 13
How many proteins have adj. P <= 0.05 according to DEqMS?
dim(
filter(res_DEqMS, sca.adj.pval <= maxAdjP)
)## [1] 645 13
How many proteins have adj. P <= 0.05 and Log2 FC >= log2(1.3) [30%] according to DEqMS?
dim(res_DEqMS %>%
filter(sca.adj.pval <= maxAdjP) %>%
filter( logFC >= minLog2FC | logFC <= -1*minLog2FC ))## [1] 154 13
gr1 <- strsplit(current_contrast, "-")[[1]][[2]]
gr2 <- strsplit(current_contrast, "-")[[1]][[1]]
#Add categorical annotation columns for LIMMA and DEqMS
res_DEqMS$Diff_LIMMA <- apply(
res_DEqMS, 1, function(x) {
local_p <- as.numeric(x[["adj.P.Val"]])
local_fc <- as.numeric(x[["logFC"]])
if (local_p <= maxAdjP & local_fc >= minLog2FC) {
return( paste("Up in", gr2) )
} else if (local_p <= maxAdjP & local_fc <= -1*minLog2FC) {
return( paste("Up in", gr1) )
} else {
return('Non-significant')
}
}
)
res_DEqMS$Diff_DEqMS <- apply(
res_DEqMS, 1, function(x) {
local_p <- as.numeric(x[["sca.adj.pval"]])
local_fc <- as.numeric(x[["logFC"]])
if (local_p <= maxAdjP & local_fc >= minLog2FC) {
return( paste("Up in", gr2) )
} else if (local_p <= maxAdjP & local_fc <= -1*minLog2FC) {
return( paste("Up in", gr1) )
} else {
return('Non-significant')
}
}
)Volcano plot for LIMMA
ggplot(
res_DEqMS,
aes(x = logFC, y = log.adj.P.Val, colour = Diff_LIMMA )
)
geom_point(shape=19, size=2, alpha = 0.6)
geom_hline(yintercept = -1*log10(maxAdjP), colour = "gray65")
geom_vline(xintercept = 0, colour = "gray65")
geom_vline(xintercept = -1*minLog2FC, colour = "gray65")
geom_vline(xintercept = minLog2FC, colour = "gray65")
ggtitle(
paste(
"LIMMA ", current_contrast,
" Adjusted P-value<=", maxAdjP, " Log2 FC>=", minLog2FC,
sep=""
)
)
theme_classic()
theme(
legend.title = element_blank(), legend.text = element_text(size=12),
plot.title = element_text(size=16)
)
labs(x = paste("Log2 FC",current_contrast), y = "-Log10 Adj. P-value" )
geom_text(
data = subset(res_DEqMS, logFC >=1.0 | logFC <= -0.8),
aes( logFC, log.adj.P.Val, label = Gene),
alpha = 0.6, hjust = 0.5, vjust = -0.6
)
Volcano plot for DEqMS
ggplot(
res_DEqMS,
aes(x = logFC, y = log.sca.adj.pval, colour = Diff_DEqMS )
)
geom_point(shape=19, size=2, alpha = 0.6)
geom_hline(yintercept = -1*log10(maxAdjP), colour = "gray65")
geom_vline(xintercept = 0, colour = "gray65")
geom_vline(xintercept = -1*minLog2FC, colour = "gray65")
geom_vline(xintercept = minLog2FC, colour = "gray65")
ggtitle(
paste(
"DEqMS ", current_contrast,
" Adjusted P-value<=", maxAdjP, " Log2 FC>=", minLog2FC,
sep=""
)
)
theme_classic()
theme(
legend.title = element_blank(), legend.text = element_text(size=12),
plot.title = element_text(size=16)
)
labs(x = paste("Log2 FC",current_contrast), y = "-Log10 Adj. P-value" )
geom_text(
data = subset(res_DEqMS, logFC >=0.9 | logFC <= -0.8),
aes( logFC, log.sca.adj.pval, label = Gene),
alpha = 0.6, hjust = 0.5, vjust = -0.6
)
Furthermore, gene set enrichment using fgsea package can be employed for basic functional annotation.
I downloaded the Gene Ontology annotations for yeast from Uniprot, with gene names as identifiers. GO Consortium data is available under Creative Commons Attribution 4.0 Unported License. Let's use the GO Biological Process as an example, the gmt file can be found in this repository as well:
annotations <- gmtPathways("go-bp_gene-symbols_uniprot.gmt")
#Prepare the rank file for the ura2-met6 contrast
#Replace the UNIPROT accession with the Gene name
dfGSEA <- res_DEqMS[c("Gene", "logFC", "log.sca.adj.pval")]Use the product (Log2FC * -Log10(adj.P.val.SCA)) to rank genes in the table:
dfGSEA$rank <- apply(
dfGSEA, 1, function(x) {
as.numeric(x[[2]]) * as.numeric(x[[3]])
}
)
ranks <- as.numeric(dfGSEA$rank)
names(ranks) <- dfGSEA$Gene
ranks <- sort(ranks)
ranks[1:10]## HIS4 RTN2 SIP18 HSP12 GCV2 SHM2 HSP26 SAM1 SAH1 GPH1
## -39.522063 -10.911242 -8.174439 -7.404254 -6.980468 -6.839459 -5.945264 -5.255033 -4.994689 -4.869912
Run fgsea and go with the default settings for enrichment:
res_gsea <- fgseaMultilevel(annotations, ranks, minSize=15, maxSize=500)
head(res_gsea)## pathway pval padj log2err ES NES size
## 1: DNA repair [GO:0006281] 0.5493827 0.9792398 0.07569463 0.4890924 0.9355784 20
## 2: actin filament organization [GO:0007015] 0.9277344 0.9939024 0.04802204 -0.3478877 -0.6806408 21
## 3: aerobic respiration [GO:0009060] 0.4152893 0.9675056 0.09110731 0.5546527 1.0388591 18
## 4: carbohydrate metabolic process [GO:0005975] 0.7070312 0.9792398 0.06064040 -0.4277279 -0.8368478 21
## 5: cell cycle [GO:0007049] 0.8966942 0.9939024 0.05216303 0.3572035 0.6690387 18
## 6: cell division [GO:0051301] 0.2551440 0.9181034 0.12154328 0.6158352 1.1780231 20
## leadingEdge
## 1: HRT1,BDF1,HTA2,RFA2,MGM101,MLP1,...
## 2: ENT2,SAC6,RHO1
## 3: RIB1,MIX14,RIB3,QCR7,MIX17,RIP1
## 4: PGM2,CIT1,GPD1,GLC3,TAL1,MDH1,...
## 5: CDC25,BNI5,NSP1
## 6: HRT1,CDC25,UBC9
Ten pathways with the largest positive enrichment scores:
res_gsea[ES > 0][head(order(pval), n=10),]## pathway pval
## 1: maturation of LSU-rRNA from tricistronic rRNA transcript (SSU-rRNA, 5.8S rRNA, LSU-rRNA) [GO:0000463] 0.05785124
## 2: chromatin remodeling [GO:0006338] 0.06918239
## 3: ribosomal large subunit biogenesis [GO:0042273] 0.07171315
## 4: fungal-type cell wall organization [GO:0031505] 0.07370518
## 5: transcription by RNA polymerase II [GO:0006366] 0.08176101
## 6: negative regulation of transcription by RNA polymerase II [GO:0000122] 0.10062893
## 7: regulation of transcription by RNA polymerase II [GO:0006357] 0.17283951
## 8: ribosome biogenesis [GO:0042254] 0.19591837
## 9: positive regulation of transcription by RNA polymerase II [GO:0048634] 0.24696356
## 10: cell division [GO:0051301] 0.25514403
## padj log2err ES NES size leadingEdge
## 1: 0.4832041 0.2712886 0.7913647 1.482218 18 NOP2,NOP12,CIC1,TIF6,RPF2
## 2: 0.4832041 0.2489111 0.7765741 1.440644 17 CHZ1,CBF1,HTZ1,TAF14,BDF1,CYC8,...
## 3: 0.4832041 0.2377938 0.7208805 1.497722 30 NOP2,RPL26A,RPL26B,TIF6,MAK21,REH1,...
## 4: 0.4832041 0.2343926 0.6997807 1.462006 31 CIS3,ZEO1,BAR1,BGL2,EXG1,CRH1,...
## 5: 0.4832041 0.2279872 0.7579933 1.400889 16 RPC10,RPB9,TAF14,BUR6,TFA1,RPB3,...
## 6: 0.5132075 0.2042948 0.7395614 1.366824 16 PAF1,RPA12,CBF1,BUR6,ESS1,RAP1,...
## 7: 0.8013468 0.1511488 0.6526132 1.248375 20 PAF1,CBF1,SPN1,HTZ1,TUP1,CYC8,...
## 8: 0.8326531 0.1404062 0.6386159 1.241881 21 RPC10,RPA14,RPA12,RPL40A,RPC19
## 9: 0.9181034 0.1226792 0.5842056 1.187375 26 KAE1,CBF1,BUR6,ESS1,RAP1,TUP1,...
## 10: 0.9181034 0.1215433 0.6158352 1.178023 20 HRT1,CDC25,UBC9
Ten pathways with the largest negative enrichment scores:
res_gsea[ES < 0][head(order(pval), n=10),]## pathway pval padj log2err
## 1: purine nucleotide biosynthetic process [GO:0006164] 0.006649502 0.3391246 0.40701792
## 2: cellular response to oxidative stress [GO:0034599] 0.040935673 0.4832041 0.31532483
## 3: ergosterol biosynthetic process [GO:0006696] 0.041904762 0.4832041 0.30775005
## 4: glycolytic process [GO:0006096] 0.085271318 0.4832041 0.21392786
## 5: retrograde vesicle-mediated transport, Golgi to endoplasmic reticulum [GO:0006890] 0.535019455 0.9792398 0.07417590
## 6: endocytosis [GO:0006897] 0.543859649 0.9792398 0.07343814
## 7: translation [GO:0006412] 0.609523810 0.9792398 0.06658921
## 8: ribosomal large subunit assembly [GO:0000027] 0.705426357 0.9792398 0.06037864
## 9: carbohydrate metabolic process [GO:0005975] 0.707031250 0.9792398 0.06064040
## 10: rRNA processing [GO:0006364] 0.727626459 0.9792398 0.05909548
## ES NES size leadingEdge
## 1: -0.8517795 -1.5621756 15 MTD1,ADE17,ADE4,ADE2,ADE6,ADE8,...
## 2: -0.6998521 -1.5312424 39 HSP12,ACT1,TPS1,GAD1,YHB1,MCR1,...
## 3: -0.7858271 -1.4633302 16 ERG27,ERG13,MCR1,ERG20,ERG9
## 4: -0.7048514 -1.3727129 20 HXK1,GLK1,CDC19,TDH1,PFK2,PGI1,...
## 5: -0.5006317 -0.9619512 19 RER1,YPT1,SEC21,COP1,GET3,ERV46
## 6: -0.4312358 -0.9435230 39 ACT1,CHC1,ENT2,DNF1
## 7: -0.4857894 -0.9046142 16 RPL24B
## 8: -0.4277358 -0.8330244 20 RPL24B,RPL3
## 9: -0.4277279 -0.8368478 21 PGM2,CIT1,GPD1,GLC3,TAL1,MDH1,...
Let's look at one of the pathways from the top with positive enrichment scores:
plotEnrichment(
annotations[["ribosomal large subunit biogenesis [GO:0042273]"]],
ranks
) labs(title="ribosomal large subunit biogenesis [GO:0042273]")
annotations[["ribosomal large subunit biogenesis [GO:0042273]"]]## [1] "ACL4" "AFG2" "ALB1" "EBP2" "ERB1" "HAS1" "HRR25" "JIP5" "JJJ1" "LOC1" "MAK11" "MAK16"
## [13] "MAK21" "MAK5" "MRT4" "NIP7" "NOC2" "NOC3" "NOG1" "NOP15" "NOP16" "NOP2" "NOP4" "NOP7"
## [25] "NOP8" "NSA1" "NSA2" "PRP43" "PUF6" "REH1" "REI1" "RIX7" "RLI1" "RLP24" "RLP7" "RPL14A"
## [37] "RPL14B" "RPL26A" "RPL26B" "RPL33A" "RPL33B" "RPL7A" "RPL7B" "RRP14" "RRP8" "RRS1" "SDA1" "SQT1"
## [49] "SYO1" "TIF4631" "TIF6" "TRM112" "YTM1"
And at one of the pathways from the top with negative enrichment scores:
plotEnrichment(
annotations[["purine nucleotide biosynthetic process [GO:0006164]"]],
ranks
) labs(title="purine nucleotide biosynthetic process [GO:0006164]")
annotations[["purine nucleotide biosynthetic process [GO:0006164]"]]## [1] "ADE1" "ADE12" "ADE13" "ADE16" "ADE17" "ADE2" "ADE3" "ADE4" "ADE5,7" "ADE6" "ADE8" "MIS1" "MTD1"
## [14] "PRS1" "PRS2" "PRS3" "PRS4" "PRS5"