Error in solve.default(t(mod) %*% mod) : Lapack routine dgesv: system is exactly singular: U[4,4] = 0
0
0
Entering edit mode
aina.jene ▴ 10
@ainajene-12984
Last seen 4.3 years ago
ISGlobal - Barcelona

Hello,

I have recently gotten this error message:

Error in solve.default(t(mod) %*% mod) :  Lapack routine dgesv: system is exactly singular: U[4,4] = 0

when trying to estimate the number of surrogate variables.

 

My data is the next one:

                                      Call Status  race           id
TCGA-2A-A8VL-01A normal  tumor white TCGA-2A-A8VL
TCGA-2A-A8VO-01A normal  tumor white TCGA-2A-A8VO
TCGA-2A-A8VX-01A normal  tumor white TCGA-2A-A8VX
TCGA-2A-A8W3-01A normal  tumor white TCGA-2A-A8W3
TCGA-2A-AAYF-01A normal  tumor white TCGA-2A-AAYF
TCGA-2A-AAYO-01A normal  tumor white TCGA-2A-AAYO

 

Levels Call: Normal, LOY, XYY

Levels Status: Tumor, Normal

Race: I am selecting just white individuals

 

And the expression data is:

            TCGA-2A-A8VL-01A TCGA-2A-A8VO-01A TCGA-2A-A8VT-01A TCGA-2A-A8VV-01A
?|100133144               21               20               49               22
?|100134869               18               14               60               17
?|10357                        81               63               95              109
?|10431                    1312              865              583              919
?|155060                    316              379              730              238
?|340602                       1                1                1                3
           

My script is the next one:

ii <- intersect(colnames(counts.f), rownames(pheno))

pheno <- pheno[ii,]
counts <- counts.f[,ii]

mod <- model.matrix(~Call + Status, data=pheno[ii,])
mod0 <- model.matrix(~1, data=pheno[ii,])

library(sva)
ns <- num.sv(counts.f[,ii], mod, method="be")
svobj <- sva(counts.f[,ii], mod, mod0, n.sv=ns)
modSv <- cbind(mod, svobj$sv)

 

When I try num.sv function in most of the datasets that I have I get the same error.

I have found these post from few years ago: ComBat_ Error in solve.default(t(design) %*% design): Lapack routine dgesv: system is exactly singular: U[4, 4] = 0

But for what I understand they already know the batch, so they are using the ComBat function. In my case, I am using sva function to estimate batch and other artifacts.

 

Can you help me find out how can I fix it?

Thank you

 

Aina Jene

 

 

sva R removebatcheffect • 1.4k views
ADD COMMENT
0
Entering edit mode

Hi Aina, did you find a solution yet? I get the very same error....

Thanks!
Sebastian

 

ADD REPLY

Login before adding your answer.

Traffic: 382 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6