Question: Deseq2: 3 factors design
0
gravatar for marburg2107
6 months ago by
marburg21070 wrote:

It's really complicated for me to handle three-factors with DESeq2, though many of the posts on Biostars.I think the developer should make it more clear to users.

Three factors are temperature(A, B); water content (C, D); co2 concentration (E, F).

I want to know the following effects: (1)temperature, (2)water, (3)co2, (4)temperatureco2, (5)waterco2, (6)temperaturewaterco2, (7)temperature*water.

I have tried a lot with design in Deseq2, like DESeqDataSetFromMatrix(countData = countData,colData = colData,design = ~co2+temperature+water+co2:temperature+co2:water)...but no design can tell me temperaturewaterco2 effect.

Thanks!

three factors design deseq2 • 395 views
ADD COMMENTlink modified 6 months ago • written 6 months ago by marburg21070
0
gravatar for marburg2107
6 months ago by
marburg21070 wrote:

DESeqDataSetFromMatrix(countData = countData,colData = colData,design = ~co2+warm+drought+co2:warm+co2:drought+warm:drought+co2:warm:drought) --------------Error in checkFullRank(modelMatrix)

My R version is 3.4.2 (2017-09-28), Deseq2 version is 1.18.1

ADD COMMENTlink written 6 months ago by marburg21070

could you post the contents of your design matrix

ADD REPLYlink written 6 months ago by russhh3.5k

Total 72 samples, 6 treatment,12 replicates.

6 treatments were: co2-400ppm, co2-400ppm+drought, co2-400ppm+drought+warm, co2-600ppm, co2-600ppm+drought,co2-600ppm+drought+warm

ADD REPLYlink modified 6 months ago • written 6 months ago by marburg21070

the actual design matrix. From the description in the first post you've got 2^3 = 8 treatments. Please. The actual design matrix

ADD REPLYlink written 6 months ago by russhh3.5k

The design of the experiment looks like:


sample co2 warm drought
a11 400ppm yes 125mm
a12 400ppm yes 125mm
a13 400ppm yes 125mm
a14 400ppm yes 125mm
a21 400ppm yes 125mm
a22 400ppm yes 125mm
a23 400ppm yes 125mm
a24 400ppm yes 125mm
a31 400ppm yes 125mm
a32 400ppm yes 125mm
a33 400ppm yes 125mm
a34 400ppm yes 125mm
e11 480ppm yes 125mm
e12 480ppm yes 125mm
e13 480ppm yes 125mm
e14 480ppm yes 125mm
e21 480ppm yes 125mm
e22 480ppm yes 125mm
e23 480ppm yes 125mm
e24 480ppm yes 125mm
e31 480ppm yes 125mm
e32 480ppm yes 125mm
e33 480ppm yes 125mm
e34 480ppm yes 125mm
A11 400ppm no 125mm
A12 400ppm no 125mm
A13 400ppm no 125mm
A14 400ppm no 125mm
A21 400ppm no 125mm
A22 400ppm no 125mm
A23 400ppm no 125mm
A24 400ppm no 125mm
A31 400ppm no 125mm
A32 400ppm no 125mm
A33 400ppm no 125mm
A34 400ppm no 125mm
E11 480ppm no 125mm
E12 480ppm no 125mm
E13 480ppm no 125mm
E14 480ppm no 125mm
E21 480ppm no 125mm
E22 480ppm no 125mm
E23 480ppm no 125mm
E24 480ppm no 125mm
E31 480ppm no 125mm
E32 480ppm no 125mm
E33 480ppm no 125mm
E34 480ppm no 125mm
aa11 400ppm no 250mm
aa12 400ppm no 250mm
aa13 400ppm no 250mm
aa14 400ppm no 250mm
aa21 400ppm no 250mm
aa22 400ppm no 250mm
aa23 400ppm no 250mm
aa24 400ppm no 250mm
aa31 400ppm no 250mm
aa32 400ppm no 250mm
aa33 400ppm no 250mm
aa34 400ppm no 250mm
ee11 480ppm no 250mm
ee12 480ppm no 250mm
ee13 480ppm no 250mm
ee14 480ppm no 250mm
ee21 480ppm no 250mm
ee22 480ppm no 250mm
ee23 480ppm no 250mm
ee24 480ppm no 250mm
ee31 480ppm no 250mm
ee32 480ppm no 250mm
ee33 480ppm no 250mm
ee34 480ppm no 250mm

ADD REPLYlink modified 6 months ago • written 6 months ago by marburg21070
  1. DESeqDataSetFromMatrix(countData = countData,colData = colData,design = ~ co2 + warm+ drought+co2:warm+co2:drought) (no problem)

  2. DESeqDataSetFromMatrix(countData = countData,colData = colData,design = ~ co2 + warm+ drought+co2:warm+co2:drought+co2:warm:drought)(the model matrix is not full rank)

ADD REPLYlink written 6 months ago by marburg21070

Thanks for your help!

I just have 6 treatments as co2(400ppm),co2(480ppm), co2(400ppm)+drought(125mm), co2(480ppm)+drought(125mm),co2(400ppm)+drought(125mm)+warm(yes),co2(400ppm)+drought(125mm)+warm(no).

As you said, we should have 8 treatments. BUT they did not do other treatments.

I want to know the following effects: (1)warm, (2)drought, (3)co2, (4)warm+co2, (5)drought+co2, (6)warm+drought+co2, (7)warm+drought from above 6 treatments.

ADD REPLYlink written 6 months ago by marburg21070
0
gravatar for marburg2107
6 months ago by
marburg21070 wrote:

countData = read.table ("C:/Users/bei/Desktop/deseq2/kegg.txt", sep= "\t", row.names = 1, head = TRUE)
colData = read.table ("C:/Users/bei/Desktop/deseq2/151617-new2.txt", sep= "\t", row.names = 1, header = TRUE)
colData[['co2']] = factor(colData[['co2']], levels = c('TRUE', 'FALSE'))
colData[['warm']] = factor(colData[['warm']], levels = c('TRUE', 'FALSE'))>
colData[['drought']] = factor(colData[['drought']], levels = c('TRUE', 'FALSE'))
dataset <- DESeqDataSetFromMatrix(countData = countData,colData = colData,design = ~ co2 + warm+ drought)
dds <- DESeq(dataset)
resultsNames(dds)
results(dds, contrast=list(c("co2_FALSE_vs_TRUE", "warm_FALSE_vs_TRUE", "drought_FALSE_vs_TRUE")))

Is it right to calculated the effect of co2droughtwarm?

Design matrix is:

sample co2 warm drougt
1 TRUE TRUE TRUE
2 TRUE FALSE FALSE
3 FALSE TRUE FALSE

...

ADD COMMENTlink modified 6 months ago • written 6 months ago by marburg21070
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1494 users visited in the last hour