Box plots with logFC values
0
0
Entering edit mode
9 weeks ago
myfam128 • 0

Hi

I am trying to draw box plots to show change in expression value between two conditions using deseq2.

d <- plotCounts(dds, gene="gene_id", intgroup="condition", returnData = TRUE)
ggplot(d, aes(x=condition, y=log2(count))) +  geom_boxplot()


Is this enough to show the differential expression of a gene between two conditions?

I would like to draw boxplots with logFC values on Y axis for a single gene in two conditions. Can please let me know how to do that?

boxplot logFC • 492 views
0
Entering edit mode

How can you do a box plot of a single LogFC value?

0
Entering edit mode

As in you would like to display the logFC as an annotation; or you would like to transform it so that every datapoint is shown not in its absolute log2(count) value; but its relative value relative to the mean of condition 1: log2(count) - mean_condition1_log2count?

0
Entering edit mode

I have seen in many research papers logFC boxplots between two conditions

example

Is this as you said above a relative value

0
Entering edit mode

In your post you're plotting the expression values of a gene, but in this figure they are plotting the log2FCs. Can you explain more what you want to show with your figure?

0
Entering edit mode

I am trying to show If a gene is up-regulated or down-regulated

0
Entering edit mode

I'm not familiar with the publication that figure is drawn from; but one way to define a per-sample logFC value is to measure the fold change relative to one of the group means. So...as i said above, just have something like cond1_mean = mean(log2(subset(data, gene=gene_id & condition=cond1)\$expr)); ggplot(subset(data=gene_id)) + geom_boxplot(aes(x=condition, y=log2(expr)-cond1_mean))

0
Entering edit mode

Thank you I will try that