nnNorm is an algorithm designed to remove intensity and spatial dependent biases from microarray expression data sets. It uses robust neural networks to fit the log-ratios on spatial coordinates and log-average intensity of the cDNA spots.
The nnNorm R package is available from the Bioconductor website here.
Details about the method can be found in the publication below:
- Tarca AL, Cooke JE, Mackay J. A robust neural networks approach for spatial and intensity-dependent normalization of cDNA microarray data. Bioinformatics 2005;21(11):2674-83.