Hello, I would like to know how exactly RMA does with the PM data. I've read 'Exploration, Normalization,...' by Irizarry et al., but for the life of me could not figure out how exactly the expression value is being calculated (especially the B(.) part - what's actually being done??). Looking at the source code should be the very last resort - since it never explicitly states RMA in the remarks, it's a little hairy as of now. It's nice that it works...but does anybody know a little more detailed description of this mysterious algorithm? Could someone explain? Thanks! Shin Wachi