Question: DESeq2 design matrix choice
0
gravatar for Yang WY
19 days ago by
Yang WY0
China Beijing
Yang WY0 wrote:

Hi!

I've downloaded HTSeq-Counts of 48 patients from TCGA, all of the patients have a NT aliquot and a TP aliquot,I want to find the DEGs of two conditions using DESeq2,

myfactor is:

> head(myfactor)
  patient tissue
1       1     NT
2       2     NT
3       3     NT
4       4     NT
5       5     NT
6       6     NT
> tail(myfactor)
   patient tissue
91      43     TP
92      44     TP
93      45     TP
94      46     TP
95      47     TP
96      48     TP

now I have alternative in design of DESeqDataSetFromMatrix().

DEA_DESeq<-DESeqDataSetFromMatrix(countData = count, colData = myfactor,design=~tissue)  #1
DEA_DESeq<-DESeqDataSetFromMatrix(countData = count, colData = myfactor,design=~patients+tissue)   #2

So which design is better?

I appreciate your help,

Yang

deseq2 design matrix • 53 views
ADD COMMENTlink modified 19 days ago by Michael Love25k • written 19 days ago by Yang WY0
Answer: DESeq2 design matrix choice
0
gravatar for Michael Love
19 days ago by
Michael Love25k
United States
Michael Love25k wrote:

If you want to find the difference in TP vs NT while controlling for patient baseline, you would provide the patient variable in the design (e.g. #2 above).

The question of what design is more about your goals in the analysis. I'm just guessing based on your mentioning the patient ID that you likely want to control for patient baseline, but these choices are up to you as the analyst. The support site is more for questions about the software and how to use it, and not intended for statistical consultation about what is the appropriate design for a given dataset.

ADD COMMENTlink written 19 days ago by Michael Love25k

Thanks for your advice! Since the clinial data varies in cases, I use patients id to represent the difference among patients. As you said, I want to find the difference in TP vs NT while controlling for patient baseline for more exact results, so #2 might be better.

ADD REPLYlink written 19 days ago by Yang WY0
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 16.09
Traffic: 287 users visited in the last hour