Time-series analysis with control and treated samples, and Fold Change extraction in DESeq2
1
0
Entering edit mode
@victor-chano-20932
Last seen 13 months ago
Madrid (UPM)

Hi everyone!

I did a question some months ago about the analysis of a multifactorial RNA-Seq experiment here: https://support.bioconductor.org/p/122100/

Now, I post a different question because we finally decided to separate genotypes and local/systemic responses, so the final design is a time-series experiment with 4 sampling times (6 h, 24 h, 72 h and 144 h after infection) and two conditions, control and treated plants. The coldata is:

>print(coldata)
             plant  condition    time
A.1_C_L_6    “1”    “control”    “006”
A.3_C_L_6    “3”    “control”    “006”
A.36_C_L_6   “36”   “control”    “006”
A.47_C_L_6   “47”   “control”    “006” 
A.4_C_L_24   “4”    “control”    “024”
A.20_C_L_24  “20”   “control”    “024”
A.22_C_L_24  “22”   “control”    “024”
A.43_C_L_24  “43”   “control”    “024”
A.2_C_L_72   “2”    “control”    “072” 
A.11_C_L_72  “11”   “control”    “072” 
A.34_C_L_72  “34”   “control”    “072”
A.42_C_L_72  “42”   “control”    “072”
A.8_C_L_144  “8”    “control”    “144”
A.21_C_L_144 “21”   “control”    “144” 
A.37_C_L_144 “37”   “control”    “144”
A.39_C_L_144 “39”   “control”    “144”
A.1_C_S_6    “13”   “treatment”  “006”
A.3_C_S_6    “15”   “treatment”  “006”
A.36_C_S_6   “23”   “treatment”  “006”
A.47_C_S_6   “50”   “treatment”  “006”
A.4_C_S_24   “16”   “treatment”  “024”
A.20_C_S_24  “33”   “treatment”  “024”
A.22_C_S_24  “41”   “treatment”  “024”
A.43_C_S_24  “44”   “treatment”  “024”
A.2_C_S_72   “18”   “treatment”  “072”
A.11_C_S_72  “31”   “treatment”  “072”
A.34_C_S_72  “46”   “treatment”  “072”
A.42_C_S_72  “49”   “treatment”  “072”
A.8_C_S_144  “9”    “treatment”  “144”
A.21_C_S_144 “12”   “treatment”  “144”
A.37_C_S_144 “24”   “treatment”  “144”
A.39_C_S_144 “45”   “treatment”  “144”

I want to identify those genes differentially expressed between conditions but also along time, so the full and the reduced models are:

design(dds) <- formula(~ time + condition + time:condition)
dds <- DESeq(dds, test="LRT", reduced = ~ time)

With this design I expect to generate p-values for differences between conditions for all time points and differences in the effect of the inoculation at different times. Is this design right?

I would also like to extract FC values between conditions for each time. Using resultsNames() I have this:

resultsNames(dds)
[1] "Intercept"                      "time_024_vs_006"               
[3] "time_072_vs_006"                "time_144_vs_006"               
[5] "condition_treatment_vs_control" "time024.conditiontreatment"    
[7] "time072.conditiontreatment"     "time144.conditiontreatment"

I have being reading that the argument "name" is used to extract fold changes, but I don't know in this case which name should I use to extract FC of the comparisons between control and inoculated for each time.

Thank you so much. Víctor

deseq2 time-serie time-course fold-change • 182 views
ADD COMMENT
1
Entering edit mode
swbarnes2 ▴ 890
@swbarnes2-14086
Last seen 28 minutes ago
San Diego

I would also like to extract FC values between conditions for each time.

The simple way to do this is to make a new col of colData which is time and condition combined. Make that column your design, and use contrast to compare subsets to each other. When you use contrast, it doesn't matter if the exact comparison you want is in ResultNames.

http://bioconductor.org/packages/release/bioc/vignettes/DESeq2/inst/doc/DESeq2.html#interactions

ADD COMMENT
0
Entering edit mode

Thank you for your answer. When you say "make that column your design" do you mean to use this new variable for the full model design?

design(dds) <- formula(~ new_variable)
dds <- DESeq(dds, test="LRT", reduced = ~ 1)
ADD REPLY

Login before adding your answer.

Traffic: 232 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6