GCRMA

The GCRMA normalisation routine performs a quantile normalisation [1] with the help of probe sequence.

DNA sequences are made up of four bases (nucleotides), called A, C, G and T. In double-stranded DNA, A is always opposite T and C is always opposite G. Each G-C pair has three hydrogen bonds between the G and the C, whereas each A-T pair has only two hydrogen bonds between the A and the T. It is for this reason that biologists are interested in the ``GC content'' of certain areas of the genome. RNA is single stranded and is generally created by transcribing (copying) DNA, with the only difference in sequence being that T is replaced by U. In microarray hybridisations, the probes printed on the chip are single-stranded cDNA (reverse-transcribed from RNA), but the target RNA contains the bases A, C, G and U, rather than A, C, G and T, so it is possible to get both A-T pairs and A-U pairs in the hybridisations. The A-U pairs have the same properties as the A-T pairs in double-stranded DNA, i.e. only two hydrogen bonds, so the idea of ``GC content'' is just as relevant to microarray hybridisations as it is to double-stranded DNA molecules. GCRMA routine takes GC content into account when doing RMA normalisation [2].

Bjørn Kåre Alsberg 2006-04-06