Quality control for chips

Go to parent Introduction to R/Bioconductor for analysis of microarray data#Tutorials

Background

RNA is a fickle molecule, and despite impressive improvements over the last decade, thing can still go spectacularly wrong along the lab pipeline. If this happens, it's quite possible that a complete chip may be bad, in the sense that it contains mostly or largely data that does not represent the biological reality.

Bad chips can obviously have disastrous results on the data analysis, either hiding biological signal or leading to false conclusions. A good data analysis pipeline should therefore include some quality control on the chips.

In practice, numerous diagnostical tools, usually of a graphical nature, have been proposed: from reconstructed images of the chip to detect spatial artifacts to standardized aggregated measures of variability.

Vignette?

==