Microarray based research

Microarrays enable the abundance of up to 50,000 genes or transcripts to be estimated in a single experiment. Microarrays have been widely applied to a broad variety of organisms for classification purposes and to identify genes or transcripts that are differentially expressed between two samples, for example diseased and healthy cell lines.

Compared to other platforms results obtained from microarrays are thought to be less reproducible; rigid grid fitting and spot finding algorithms, which fail to account for the wide variation in experimental images, are known to contribute to this problem. Attempts to reduce experimental variability divert scarce resources into higher cost consumables, equipment and replicated experiments. Operator intervention becomes a major source of experimental variability, progress towards major research objectives is slowed and confidence is further undermined by the arbitrary way in which results may be flagged for exclusion. There is a widespread belief that these problems are insoluble and that low confidence has to be attributed to all but the most compelling results.

BlueFuse automates the analysis of microarray experiments by using advanced statistical modelling technology to 'learn' the characteristics of signal and noise from the microarray image.

read more

© cambridge bluegnome 2006 info@cambridgebluegnome.com