Low-pass frequency-domain filtering of oligonucleotide microarray data images

Oligonudeotide microarrays technologies offer the possibility of simultaneously monitoring thousands of hybridization reactions. These arrays show high potential for many medical and scientific applications as gene expression monitoring, sequence analysis, and genotyping. Nevertheless microarrays are exposed to errors during manufacturing, similar to silicon circuit electronics and the hybridization process may be contaminated by different reasons. Other source of errors is due to optical noise during scanning and processing, or to interactions between molecular structures and light (dispersion among others). To reduce some of these effects are used replicates of experiments with the cost of increasing expenses. In order to detect noise contamination in microarray data images well-known computational techniques are proposed to help in visual analysis. Some experiments are shown to check the effectiveness of these approaches.