Hi,
I am working on differential splicing caused by 4 treatments, each with 6 replicates.
From differential expression analysis, I found that the RNA-seq data has batch effect. Therefore, I want to used the package MOCCASIN to remove the batch effect. But I have some problem with the model matrix. The model matrix I intended to use is as follow (T for treatment):
filename T1 T2 T3 T4 batch1 batch2 batch3 batch4 batch5 batch6
Sample1 1 0 0 0 1 0 0 0 0 0
Sample2 1 0 0 0 0 1 0 0 0 0
Sample3 1 0 0 0 0 0 1 0 0 0
Sample4 1 0 0 0 0 0 0 1 0 0
Sample5 1 0 0 0 0 0 0 0 1 0
Sample6 1 0 0 0 0 0 0 0 0 1
Sample7 0 1 0 0 1 0 0 0 0 0
Sample8 0 1 0 0 0 1 0 0 0 0
Sample9 0 1 0 0 0 0 1 0 0 0
Sample10 0 1 0 0 0 0 0 1 0 0
Sample11 0 1 0 0 0 0 0 0 1 0
Sample12 0 1 0 0 0 0 0 0 0 1
Sample13 0 0 1 0 1 0 0 0 0 0
Sample14 0 0 1 0 0 1 0 0 0 0
Sample15 0 0 1 0 0 0 1 0 0 0
Sample16 0 0 1 0 0 0 0 1 0 0
Sample17 0 0 1 0 0 0 0 0 1 0
Sample18 0 0 1 0 0 0 0 0 0 1
Sample19 0 0 0 1 1 0 0 0 0 0
Sample20 0 0 0 1 0 1 0 0 0 0
Sample21 0 0 0 1 0 0 1 0 0 0
Sample22 0 0 0 1 0 0 0 1 0 0
Sample23 0 0 0 1 0 0 0 0 1 0
Sample24 0 0 0 1 0 0 0 0 0 1
And the command I used was
python moccasin.py modelmat.tsv ../majiq_build AfterMOCCASIN_majiq_file batch1 1 batch2 batch3 batch4 batch5 batch6 -J 8
I got the error saying my model matrix is not full rank. But how should I design a full rank model matrix to remove the batch effect.
Thanks,
Maoting