[Bioc-sig-seq] error msg from edgeR estimateCRDisp

[kristen.dang at syngenta.com]

Dear Gordon,
Many thanks for your reply.

I apologize, but I cannot find where to download version 2.0.3 of edgeR. I can only find 2.0.2 on the bioconductor site. Would you point me to the appropriate location?

Thank you for suggesting the rounding -- I had forgotten that the count data provided to me included decimals. However, after adding the rounding, I get a new error from the estimateCRDisp function:

> data = readDGE(targets)
> data$counts = round(data$counts)
> data = calcNormFactors(data)
> contrasts(targets$factor_a) = contr.sum(2)
> contrasts(targets$factor_b) = contr.sum(2)
> design = model.matrix(~factor_a + factor_b, data = targets)
> data.cd = estimateCRDisp(data, design)
Error in chol.default(crossprod(design, .vecmat(mu[i, ]/(1 + dispersion *  :
  the leading minor of order 4 is not positive definite

> traceback()
5: .Call("La_chol", as.matrix(x), PACKAGE = "base")
4: chol.default(crossprod(design, .vecmat(mu[i, ]/(1 + dispersion *
       mu[i, ]), design)))
3: chol(crossprod(design, .vecmat(mu[i, ]/(1 + dispersion * mu[i,
       ]), design)))
2: adjustedProfileLik(spline.disp[i], y.filt, design = design, offset = offset.mat.filt +
       lib.size.mat.filt)
1: estimateCRDisp(data, design)
>

Any suggestions are much appreciated.

Regards,
Kristen Dang



message may contain confidential information. If you are not the designated recipient, please notify the sender immediately, and delete the original and any copies. Any use of the message by you is prohibited.