How to import RSEM expected_count data in ucsc to a deseq2 object ?

Question

[Yun]

I have downloaded gene expression data frome https://xenabrowser.net/datapages/?dataset=TcgaTargetGtex_gene_expected_count&host=https%3A%2F%2Ftoil.xenahubs.net&removeHub=https%3A%2F%2Fxena.treehouse.gi.ucsc.edu%3A443,

I find the unit of the data is log2(expected_count+1),and I want to use these do differentil expression analysis through DESeq2 in R, So I make a transformation of (2^data -1), and want to deliver to DESeq2, there are some viewpoint I found in forum that the data directly rounded may cause some errors, so I want to use tximport to transform these data to a DESeq2 object, but it seems something wrong. here is my code and the warning:

tximport::tximport("prostate_rna", type = "rsem", txIn = F, txOut = F) reading in files with read_tsv 1 Error in computeRsemGeneLevel(files, importer, geneIdCol, abundanceCol, : all(c(geneIdCol, abundanceCol, lengthCol) %in% names(raw)) is not TRUE 此外: Warning message: Unnamedcol_typesshould have the same length ascol_names. Using smaller of the two.

[Kristoffer Vitting-Seerup]

DESeq2 requires integers so it does not matter how you import the data it will need to be rounded. Just go with the 2^data -1 transformation.

[Yun]

Thanks for your recommendation, I find another way in rsem official site-"ebseq" package, I'll try the both.