DEXSeq help - error: model matrix not full rank

0

Entering edit mode

greenhawmatt • 0

@greenhawmatt-12843

Last seen 7.7 years ago

Hi all,

I am new to using DEXSeq and I could use some help from someone familiar with it.

I am following the vignette with my dataset in stepwise fashion. My error occurs during this step "dxd = estimateDispersions( dxd )". The error message suggests using "ddx <- estimateDispersionsGeneEst(ddx)". After trying that I get the error message saying that model matrix is not full rank.

My experiment table looks like this:

	condition
S1	disease
S2	healthy

And my design formula for DEXSeqDataSetFromHTSeq is: design = ~ sample+exon + condition:exon

Any help is much appreciate. Thank you.

Best,

Matthew

dexseq • 1.2k views

ADD COMMENT • link 7.6 years ago greenhawmatt • 0

0

Entering edit mode

Hi Matthew. The error is being produced because the experiment does not seem to have replicates. Sorry, DEXSeq does not support such cases.

Alejandro

ADD REPLY • link 7.6 years ago Alejandro Reyes ★ 1.9k

0

Entering edit mode

Hi Alejandro,

Thanks for your reply. I tried running DEXSeq with replicates but I still get the same error about a non full rank matrix. Can you help with this?

<caption>matrix before "estimateDispersionsGeneEst"</caption>

sample	condition	exon	sizeFactor
<factor>	<factor>	<factor>	<numeric>
S7_1	disease	this	1
S7_2	disease	this	1
S8_1	healthy	this	1
S8_2	healthy	this	1
S7_1	disease	others	1
S7_2	disease	others	1
S8_1	healthy	others	1
S8_2	healthy	others	1

ADD REPLY • link 7.6 years ago greenhawmatt • 0

0

Entering edit mode

Would you include the full code that generates the error and the output of your sessionInfo() ?

ADD REPLY • link 7.6 years ago Alejandro Reyes ★ 1.9k

Login before adding your answer.