Go Back   SEQanswers > Bioinformatics > Bioinformatics

Similar Threads
Thread Thread Starter Forum Replies Last Post
DESeq2 help.. design pinuscatcus Bioinformatics 1 04-10-2015 05:53 AM
DESeq2 design error evenstar Bioinformatics 1 02-10-2015 01:09 PM
DESeq2 multifactorial design Erinyes Bioinformatics 5 09-19-2014 08:32 AM
Deseq2 design Wei-HD Bioinformatics 11 09-04-2014 11:13 AM
Paired design versus unpaired design in DESeq2 KristenC RNA Sequencing 1 05-29-2014 11:05 AM

Thread Tools
Old 04-13-2017, 10:25 AM   #1
Junior Member
Location: Ontario

Join Date: Apr 2017
Posts: 1
Default Linear Combinations in DESeq2 design


I have an RNAseq experiment where I am studying longterm effects from a pregnancy condition in mice. Although there are expression similarities between littermates (observed on my bioplot and sample-to-sample dendrograms), long-term metabolic responses within a litter vary. I would like to know what genes are changes with the pregnancy condition but regress out the changes due to different litters. Because the group is a pregnancy condition, “MNR” and controls cannot be in the same litter, resulting in two of my variables being a linear combination.

The summarized experiment is the following:
ID Group Litter_alph
C1 Control b
C3 Control c
C7 Control c
C6 Control d
C4 Control d
C8 Control d
C5 Control e
C2 Control e
I9 MNR a
I5 MNR f
F2 MNR f
T6 MNR g
F3 MNR g
T3 MNR h
T7 MNR i
T5 MNR i
I1 MNR i
T4 MNR j
T2 MNR k
I6 MNR k
F4 MNR l
I2 MNR l
I3 MNR l
I4 MNR l
I7 MNR m
T1 MNR n
F1 MNR n
T8 MNR n
I8 MNR o

Here is an example of that command:
> design(ddsMF) <- formula(~ Litter_alph + Group)

Error in checkFullRank(modelMatrix) :
the model matrix is not full rank, so the model cannot be fit as specified.
One or more variables or interaction terms in the design formula are linear
combinations of the others and must be removed.

I’ve tried to work around this by (see

> m <- model.matrix(~ID:Litter_alph + Group, se)
> m_use <- m[ , colSums(m) > 0]
> design(dds) <- ~ Group
> dds <- estimateSizeFactors(dds)
> dds <- estimateDispersionsGeneEst(dds, modelMatrix=m_use)

Error in eval(expr, envir, enclos) : inv(): matrix seems singular

Is there anyways to work around these issue and regress out litter effects? Any help would be appreciated!
R_Beth is offline   Reply With Quote

deseq2, design, linear combinations, matrix not full rank, matrix seems singular

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

All times are GMT -8. The time now is 10:41 PM.

Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2020, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO