Hi,
 

I have been working on comparing different algorithms for batch effect removal in LC-MS metabolomics data. When trying to use ComBat I encountered a couple of problems which I believe would be good if they were mentioned in the reference manual.

1. Data type for input.
There are no clear instructions on what data type is accepted as input to the function. Therefore I had to do the tutorial, check the data type given as inputs there, and then convert my data into that data type. (If I understood it correctly both batch vector and expression matrix should be "double").

2. Variance restrictions
It would be good to mention that ComBat is sensitive to low variance features and that they might need to be excluded from the dataset before using ComBat.

This information would have saved me a couple of hours of work and I am sure it would help other persons as well.