Question: champ.runCombat is failing: "invalid 'row.names' length"
0
9 months ago by
Gurkenkönig0 wrote:

I have read in an existing matrix of 450K beta values with champ.filter and linked it to a dataframe of covariates. So far everything worked perfectly (champ.QC, champ.norm, champ.SVD) only in champ.runCombat I always get the same error message.

[===========================]
[<< CHAMP.RUNCOMBAT START >>]
-----------------------------
<< Preparing files for ComBat >>
[Combat correction will be proceed with 414206 probes and 373 samples.]

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


Can anyone give me a hint what could be the problem? When checking the covariates and the beta matrix I can't see any problems at the first glance.

champ • 441 views
written 9 months ago by Gurkenkönig0

Hello:

Could you show send your pd file (Sample_Sheet.csv) to me to check what's wrong with it? To email champ450K@gmail.com.

Best

Yuan Tian

By the way, could you please check if the batch varaible (say Slides) in your pd is "Character" or "Factor" format? I found that some users' "Slides" column is numeric, which might triggle this error.

To check the format, use code: class(pd[,X]).

If it's numeric, to change the format, use code: pd[,X] = as.character(pd[,X])

Then you may retry the code.

Best

Yuan Tian