Error when running spatialdecon() on COSMx data

[hcurley]

lung12.expr <- read.delim("Lung12_exprMat_file.csv", sep=",")
rownames(lung12.expr) <- paste(lung12.expr$cell_ID, lung12.expr$fov, sep="_")
lung12.expr$cell_ID <- NULL
lung12.expr$fov <- NULL

Row.sums.expr <- rowSums(lung12.expr)
length(which(Row.sums.expr < 20))/nrow(lung12.expr)*100
#3.415062
lung12.expr2 <- lung12.expr[-which(Row.sums.expr < 20),]

lung12.expr2f <- lung12.expr2
lung12.expr2n <- lung12.expr2
for(i in 1:ncol(lung12.expr2)){
lung12.expr2f[,i] <- as.factor(lung12.expr2[,i])
lung12.expr2n[,i] <- as.numeric(as.character(lung12.expr2f[,i]))
}

Row.sums.expr2 <- rowSums(lung12.expr2n)
lung12.scaled.expr <- sweep(lung12.expr2n, 1, Row.sums.expr2, "/")

negnames = c("NegPrb3","NegPrb5","NegPrb6","NegPrb7", "NegPrb8","NegPrb9","NegPrb10","NegPrb11","NegPrb12","NegPrb13","NegPrb14","NegPrb15","NegPrb16","NegPrb17","NegPrb18","NegPrb19","NegPrb20","NegPrb21","NegPrb22","NegPrb23")

Lung12 <- list()
Lung12[[1]] <- t(lung12.scaled.expr)
Lung12[[2]] <- t(lung12.expr2n)
names(Lung12) <- c("normalized", "raw")

Lung12$bg <- derive_GeoMx_background(
norm = Lung12$normalized,
probepool = rep(1, nrow(Lung12$normalized)),
negnames = c("NegPrb3","NegPrb5","NegPrb6","NegPrb7", "NegPrb8","NegPrb9","NegPrb10","NegPrb11","NegPrb12","NegPrb13","NegPrb14","NegPrb15","NegPrb16","NegPrb17","NegPrb18","NegPrb19","NegPrb20","NegPrb21","NegPrb22","NegPrb23")
)

res0 <- spatialdecon(
norm = Lung12$normalized,
bg = Lung12$bg,
X = safeTME
)

Error in solve.default(fit$hessian) :
Lapack routine dgesv: system is exactly singular: U[1,1] = 0
In addition: Warning message:
In lognlm.fit3(X = X, y = Y, par = par0, lik = lik, opt = opt, offset = offset, :
Unsuccessful convergence

Any idea what is going on here?

[maddygriz]

Hi @hcurley,

Our customer support team will be able to help you with this. Please email geomxsupport@nanostring.com with your issue.

Thanks,
Maddy