DESeqDataSetFromMatrix error: invalid 'row.names'

Question

[jschiant]

When I'm trying to use the DESeqDataSetFromMatrix function, I'm getting the error:

Error in .rowNamesDF<-(x, value = value) : invalid 'row.names' length

Yesterday I successfully ran the function with the exact same file inputs with no problem, but somehow without changing the files at all it has stopped working. I tried restarting R, I checked to make sure the sample names lined up, I tried updating all of my packages. Why is this happening? Below is my code.

countData <- read_csv('thyroid_case_counts.csv')

metaData <- read.csv('rna_clinical_data.tsv', header = TRUE, sep = "")

dds <- DESeqDataSetFromMatrix(countData=countData, colData=metaData, design=~Cluster, tidy = TRUE)

[rpolicastro]

can you post the results of dput(head(countData)) and dput(head(metaData)) here? It's hard to guess what's going wrong without seeing a part of your data.

[jschiant]

dput(countData[1:10, 1:10])
Error: Can't subset columns that don't exist.
x Locations 9 and 10 don't exist.
ℹ There are only 8 columns.

[jschiant]

If I try to convert my first column with the gene names in it to the row names, when I run the dds line it says that my row names have duplicates, because it tries to use the counts of the first samples as the row names

[swbarnes2]

So.your count data has 8 columns, and metadata data has information for 6...?

[jschiant]

Ahhh thank you I see what the problem is, the files were formatted correctly but I must have accidentally added in a space without realizing it to the line that converted the gene ID column into the row names. Now it works, thank you!!

[rpolicastro]

For the count data you can just do dput(head(countData)) then and edit your post.