[BioC] DESeq: multi-factors testing questions

[Yanzhu [guest]]

Dear Community,

I have some questions about how the DESeq r package  works for multi-factors expersiment. My experiment has three factors: A/B/C, and 8 replicates per condition. I would like the test the significance of the main effects of factor A, B and C, the significance of the two-way interaction terms: A:B, A:C and B:C, and the significance of the three-way interaction term: A:B:C. I want the table of pvalue for each term (main effects, two-way interaction terms and the three-way interaction term) like what ANOVA does for each gene. 


I know to test the significance of the three-way interaction term, we use the following coding:
fitDeSeq1<-fitNbinomGLMs(cdsFull,count~A+B+C+A:B+A:C+B:C+A:B:C)
fitDeSeq0<-fitNbinomGLMs(cdsFull,count~A+B+C+A:B+A:C+B:C)
modDeSeq<-nbinomGLMTest(fitDeSeq1,fitDeSeq0)
                               
My Questions are: how can I test the significance of main effects and the two-way interaction terms?

1. To test the main effect of A, B and C

(i) To test the main effect of A:
fitDeSeq1<-fitNbinomGLMs(cdsFull,count~A)
fitDeSeq0<-fitNbinomGLMs(cdsFull,count~1)
modDeSeq<-nbinomGLMTest(fitDeSeq1,fitDeSeq0)

(ii) To test the main effect of B:
Do I need to use the following coding:
fitDeSeq1<-fitNbinomGLMs(cdsFull,count~B)
fitDeSeq0<-fitNbinomGLMs(cdsFull,count~1)
modDeSeq<-nbinomGLMTest(fitDeSeq1,fitDeSeq0)

OR:

Do I need to use the following coding:
fitDeSeq1<-fitNbinomGLMs(cdsFull,count~A+B)
itDeSeq0<-fitNbinomGLMs(cdsFull,count~B)
modDeSeq<-nbinomGLMTest(fitDeSeq1,fitDeSeq0)

Which one is correct?

(iii) To test the main effect of C:
Do I need to use the following coding:
fitDeSeq1<-fitNbinomGLMs(cdsFull,count~C)
fitDeSeq0<-fitNbinomGLMs(cdsFull,count~1)
modDeSeq<-nbinomGLMTest(fitDeSeq1,fitDeSeq0)

OR:

Do I need to use the following coding:
fitDeSeq1<-fitNbinomGLMs(cdsFull,count~A+B+C)
itDeSeq0<-fitNbinomGLMs(cdsFull,count~A+B)
modDeSeq<-nbinomGLMTest(fitDeSeq1,fitDeSeq0)

which one is correct?

2. To test the two-way interaction terms: A:B, A:C and B:C

(i) To test the two-way interaction term: A:B
Do I need to use the following coding:
fitDeSeq1<-fitNbinomGLMs(cdsFull,count~A+B+C+A:B)
fitDeSeq0<-fitNbinomGLMs(cdsFull,count~A+B+C)
modDeSeq<-nbinomGLMTest(fitDeSeq1,fitDeSeq0)

Is it correct?

(ii) To test the two-way interaction term: A:C
Do I need to use the following coding:
fitDeSeq1<-fitNbinomGLMs(cdsFull,count~A+B+C+A:B+A:C)
fitDeSeq0<-fitNbinomGLMs(cdsFull,count~A+B+C+A:B)
modDeSeq<-nbinomGLMTest(fitDeSeq1,fitDeSeq0)

OR:

Do I need to use the following coding:
fitDeSeq1<-fitNbinomGLMs(cdsFull,count~A+B+C+A:C)
fitDeSeq0<-fitNbinomGLMs(cdsFull,count~A+B+C)
modDeSeq<-nbinomGLMTest(fitDeSeq1,fitDeSeq0)

Which one is correct?

(iii) To test the two-way interaction term: B:C
Do I need to use the following coding:
fitDeSeq1<-fitNbinomGLMs(cdsFull,count~A+B+C+A:B+A:C+B:C)
fitDeSeq0<-fitNbinomGLMs(cdsFull,count~A+B+C+A:B+A:C)
modDeSeq<-nbinomGLMTest(fitDeSeq1,fitDeSeq0)

OR:

Do I need to use the following coding:
fitDeSeq1<-fitNbinomGLMs(cdsFull,count~A+B+C+B:C)
fitDeSeq0<-fitNbinomGLMs(cdsFull,count~A+B+C)
modDeSeq<-nbinomGLMTest(fitDeSeq1,fitDeSeq0)

Which one is correct?

Thank you!




 

 -- output of sessionInfo(): 

 sessionInfo() 
R version 3.0.1 (2013-05-16)
Platform: x86_64-w64-mingw32/x64 (64-bit)

locale:
[1] LC_COLLATE=English_United States.1252  LC_CTYPE=English_United States.1252    LC_MONETARY=English_United States.1252
[4] LC_NUMERIC=C                           LC_TIME=English_United States.1252    

attached base packages:
[1] parallel  stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] DESeq_1.12.1       lattice_0.20-15    locfit_1.5-9.1     Biobase_2.20.1     BiocGenerics_0.6.0 edgeR_3.2.4       
[7] limma_3.16.8      

loaded via a namespace (and not attached):
 [1] annotate_1.38.0      AnnotationDbi_1.22.6 DBI_0.2-7            genefilter_1.42.0    geneplotter_1.38.0  
 [6] grid_3.0.1           IRanges_1.18.4       MASS_7.3-26          RColorBrewer_1.0-5   RSQLite_0.11.4      
[11] splines_3.0.1        stats4_3.0.1         survival_2.37-4      tools_3.0.1          XML_3.98-1.1        
[16] xtable_1.7-1 

--
Sent via the guest posting facility at bioconductor.org.