Orthogonal projections to latent structures as a strategy for microarray data normalization
2007

New Method for Normalizing Microarray Data

Sample size: 26 publication Evidence: high

Author Information

Author(s): Max Bylesjö, Daniel Eriksson, Andreas Sjödin, Stefan Jansson, Thomas Moritz, Johan Trygg

Primary Institution: Umeå University

Hypothesis

Can orthogonal projections to latent structures (OPLS) effectively normalize microarray data to remove systematic biases?

Conclusion

The OPLS methodology effectively separates biological variation from systematic biases in microarray data, improving the accuracy of results.

Supporting Evidence

  • The OPLS normalization strategy showed leading average true negative and true positive rates compared to other methods.
  • The methodology does not require prior knowledge of specific biases present in the data.
  • OPLS is applicable to both dual-channel and single-channel microarray data.

Takeaway

This study shows a new way to clean up messy data from gene experiments so we can see the real differences in genes better.

Methodology

The study used OPLS, a multivariate regression method, to normalize microarray data by identifying and removing non-correlated systematic biases.

Potential Biases

The method assumes that systematic variation is orthogonal to biological sample variation, which may not always hold true.

Limitations

The methodology may not be applicable in unsupervised analyses where group information is unavailable.

Digital Object Identifier (DOI)

10.1186/1471-2105-8-207

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication