DESeq2 comparison two cell types and two genotypes
1
0
Entering edit mode
Yosapol • 0
@40f341c9
Last seen 3 months ago
Netherlands

Dear All,

I have the samples of two cell types (Embryo and Ovule) with two genotypes (WT and MT), which indicate as condition and genotype below.

I would like to compare:

1. EmbryoWT vs EmbryoMT,
2. OvuleWT vs OvuleMT.
3. EmbryoWT vs OvuleWT
4. EmbryoMT vs OvuleMT.

I used this model to compare my samples and try to look for solutions. However, It seems to be the analysis will include the same condition or genotype together which is not our interest.

dds <- DESeqDataSetFromMatrix(countData = clean_data, colData = meta, design = ~ genotype + condition + genotype:condition)

Enter the body of text here

Code should be placed in three backticks as shown below


# include your problematic code here with any corresponding output
# please also include the results of running the following in an R session

sessionInfo( )

GeneExpressionWorkflow DESeq2 GeneExpression • 640 views
0
Entering edit mode

Has been cross-posted: https://www.biostars.org/p/9467526/

2
Entering edit mode
@mikelove
Last seen 14 hours ago
United States

This is answered in the vignette section on interactions.

0
Entering edit mode

Dear Micheal,

Thank you very much for your suggestions. This is exactly what i am looking for.

### Create DESeq2Dataset object

dds <- DESeqDataSetFromMatrix(countData = clean_data, colData = meta, design = ~ celltypes + genotype)

dds$group <- factor(paste0(dds$celltypes, dds\$genotype))

design(dds) <- ~ group

### Plot DESeq2
dds <- DESeq(dds)

### Extract information
# 1. EWT VS EMT
results(dds, name = "group_EmbryoWT_vs_EmbryoMT")

# 2. OWT VS OMT
results(dds, contrast=c("group", "OvuleWT", "OvuleMT"))


Could you please suggest how to find EmbryoWT vs OvuleWT and EmbryoMT vs OvuleMT? The helps from others also highly appreciate.

Thank you very much,

0
Entering edit mode

Use contrast with those group names? I don't follow the question.

0
Entering edit mode

I apologized for the unclear massage. Here is my group.

EmbryoMT EmbryoMT EmbryoMT EmbryoWT EmbryoWT EmbryoWT OvuleMT OvuleMT OvuleMT OvuleWT OvuleWT OvuleWT

Levels: EmbryoMT EmbryoWT OvuleMT OvuleWT

Would it be possible to extract information from All WT (EmbryoWT+OvuleWT) vs All MT (EmbryoMT+OvuleMT)

1
Entering edit mode

You can use contrast with a list, e.g. list(c("A","B"), c("C","D")) and then listValues=c(1/2,-1/2)

1
Entering edit mode

Is this different or better than using a design of ~ celltypes + genotype, and contrasting MT vs WT?

1
Entering edit mode

It is different, hard to say which is better. Above is the average effect -- simply the arithmetic average of the two LFC. Your design assumes the effect is the same in both cell types. So the dispersion would be different in these two designs, for example.

0
Entering edit mode

Hi Michael,

Thank you very much.

0
Entering edit mode

Hi Michael,

I am trying the same I believe at the moment. I have diagnosis (no cancer, previous cancer, cancer) and site(arm, hand). And I want to to the following Differential Expression:

"No cancer" "arm" against "Previous cancer" + "cancer" "arm"

I managed to do the the groups as above, but still struggling to combine "previous cancer" + "cancer"

res <-results(dds, contrast=c("group", "nocancerarm", ("previouscancerarm","cancerarm"))

Can you help me? Thank you very much!

0
Entering edit mode

Use the list style of contrast. Read over ?results

0
Entering edit mode

Thanks I read it, but it is a bit hard to understand without an example.

I tried this but is giving me errors:

list(c("previouscancerarm","cancerarm")) listValues=c(1,-1)

And then

res<- results(dds, contrast=c("group", "nocancerarm", "list"))

0
Entering edit mode

Did you know that man pages have examples? There are a bunch of them at the bottom of ?results.

0
Entering edit mode

I cannot see an example for my specific case with a list of two and grouping? Thank you for any hint.

0
Entering edit mode

There are many examples of this if you search the support site:

0
Entering edit mode

One more question to this - in the vignette it uses this formula: ~genotype + condition + genotype:condition

Wouldnt I need to write the following in above sample: design = ~ celltypes + genotype + celltype:genotype ?

0
Entering edit mode

Please adapt the code to your own experiment. It would seem that, yes, you need celltypes and genotype

0
Entering edit mode

Ok thanks because above it says: dds <- DESeqDataSetFromMatrix(countData = clean_data, colData = meta, design = ~ celltypes + genotype), missing celltype:genotype