Question

EnhancedVolcano plot with colCustom

0

Entering edit mode

Neeraja Vegesna ▴ 10

@9686cfcd

Last seen 3.7 years ago

United States

I am trying to plot a volcano plot and highlight genes from multiple lists using colCustom.

My data looks as follows with 33323 rows.

Screenshot of data

I have 11 lists, ID_W1-11, each containing a list of genes. I would like to color genes from each list a different color in the volcano plot. gene lists

So I generated keyvals.colour to use colCustom with EnhancedVolcano as follows

keyvals.colour <- ifelse(rownames(data) %in% ID_W1, "#CBA7CF",
                     ifelse(rownames(data) %in% ID_W2, "#E33092",
                            ifelse(rownames(data) %in% ID_W3, "#810081",
                                   ifelse(rownames(data) %in% ID_W4, "#662965",
                                          ifelse(rownames(data) %in% ID_W5, "#93CAEE",
                                                 ifelse(rownames(data) %in% ID_W6, "#1A58A8",
                                                        ifelse(rownames(data) %in% ID_W7, "#C19769",
                                                               ifelse(rownames(data) %in% ID_W8, "#D9E650",
                                                                      ifelse(rownames(data) %in% ID_W9, "#417652",
                                                                             ifelse(rownames(data) %in% ID_W10, "#00A94F",
                                                                                    ifelse(rownames(data) %in% ID_W11, "#A7D05C",
                                                                                           "#8F8F8F")))))))))))

keyvals.colour[is.na(keyvals.colour)] <- "#8F8F8F"
names(keyvals.colour)[keyvals.colour == "#CBA7CF"] <- 'W1'
names(keyvals.colour)[keyvals.colour == "#E33092"] <- 'W2'
names(keyvals.colour)[keyvals.colour == "#810081"] <- 'W3'
names(keyvals.colour)[keyvals.colour == "#662965"] <- 'W4'
names(keyvals.colour)[keyvals.colour == "#93CAEE"] <- 'W5'
names(keyvals.colour)[keyvals.colour == "#1A58A8"] <- 'W6'
names(keyvals.colour)[keyvals.colour == "#C19769"] <- 'W7'
names(keyvals.colour)[keyvals.colour == "#D9E650"] <- 'W8'
names(keyvals.colour)[keyvals.colour == "#417652"] <- 'W9'
names(keyvals.colour)[keyvals.colour == "#00A94F"] <- 'W10'

selected colors

EnhancedVolcano(data,
                lab = rownames(data),
                x = 't0.vs..t1.Log2.Fold.Change',
                y = 't0.vs..t1.adj..p.value',
                colCustom = keyvals.colour)

The resulting plot is not displaying all the selected colours. Enhanced_volcano_plot Any suggestions on what I am doing wrong or how to incorporate the multiple colours? Thanks!

EnhancedVolcano • 4.8k views

ADD COMMENT • link 3.7 years ago Neeraja Vegesna ▴ 10

score 0 · Answer 1 · 2021-04-14

0

Entering edit mode

Kevin Blighe ★ 4.0k

@kevin

Last seen 6 weeks ago

Republic of Ireland

Hi, I developed this package. Can you possibly try with just 3 groups, say, W1, W2, and everything else combined, and see if that works?

Does ggplot2 return any warning or message when the function is run as is?

Kevin

ADD COMMENT • link 3.7 years ago Kevin Blighe ★ 4.0k

0

Entering edit mode

Hi Kevin, When I run the function in my previous post, the only warning I get is the following..

Warning message: One or more p-values is 0. Converting to 10^-1 * current lowest non-zero p-value...

I used EnhancedVolcano to highlight only 2 sets of genes and it works well. See below..

up_path <- c(ID_W1, ID_W2, ID_W3, ID_W4)
down_path <- c(ID_W8, ID_W9, ID_W10, ID_W11)

## Print volcano plots with up genes in purple and down genes in green
keyvals.colour1 <- ifelse(
  rownames(data) %in% up_path, 'purple',
  ifelse(rownames(data) %in% down_path, 'green',
         'black'))

keyvals.colour1[is.na(keyvals.colour1)] <- 'black'
names(keyvals.colour1)[keyvals.colour1 == 'black'] <- 'NA'
names(keyvals.colour1)[keyvals.colour1 == 'purple'] <- 'UP_1,2,3,4'
names(keyvals.colour1)[keyvals.colour1 == 'green'] <- 'DOWN_8,9,10,11'

EnhancedVolcano(data,
                lab = NA,
                x = 't0.vs..t1.Log2.Fold.Change',
                y = 't0.vs..t1.adj..p.value',
                colCustom = keyvals.colour1)

plots with only 2 colors

Neeraja

ADD REPLY • link 3.7 years ago Neeraja Vegesna ▴ 10

0

Entering edit mode

Hi, that's great that it works with a lower number of groups. Can you re-check the logic flow when you are using more groups? EnhancedVolcano should technically work for an infinite number of groups.

Regarding the warning, it just means that there are p-values of 0 in your dataset. As we cannot compute -log10(0), we have to impute these as non-zero values in order to include them in the plot. If your lowest non-zero p-value is 0.0004, then these will become 0.00004

ADD REPLY • link 3.7 years ago Kevin Blighe ★ 4.0k

1

Entering edit mode

Hi Kevin,

It worked once I added names(keyvals.colour1)[keyvals.colour1 == '#8F8F8F'] <- 'NA'.

volcanoplot

Thanks a lot for your help!

Neeraja

ADD REPLY • link 3.7 years ago Neeraja Vegesna ▴ 10