Dear Supporters,

I'm using RUVnormalize package to correct a noisy dataset using negative controls. The input I'm using is a log2 expression matrix (Illumina V4 microarray) and I'm using naiveRandRUV function. Everything goes fine except that the output matrix from naiveRandRUV has very low values centered at 0 (with a lot of negative values). Am I doing something wrong or there is some option I don't know to output a normal uncentered log2 matrix (as for example does RUVseq) ?

Thank you in advance for your help

Massimo Acquaviva

Bioinformatician

San Raffaele Institute      Milan   Italy

naiveRandRUV ​outputs the dataset with numbers compressed around 0. It's just a characteristic of the method. I like to avoid using RUV unless the PCA or MDS plot indicates that it's necessary.