Reproducibility assessment of independent component analysis of expression ratios from DNA microarrays
2003

Assessing the Reproducibility of Independent Component Analysis in DNA Microarray Data

Sample size: 63 publication Evidence: moderate

Author Information

Author(s): David Philip Kreil, David J. C. MacKay

Primary Institution: University of Cambridge

Hypothesis

How stable is the independent component analysis (ICA) when applied to DNA microarray data?

Conclusion

The study found that ICA on yeast gene expression ratio data is robust, with most signatures remaining identifiable even after significant data removal.

Supporting Evidence

  • 10 different random number generator seeds were used to assess reproducibility.
  • 63 yeast wild-type vs. wild-type experiments were analyzed.
  • 10 reliably identified signatures were found, indicating that variance is not just noise.

Takeaway

Scientists used a method to analyze gene data and found that even when they removed some data, they could still recognize important patterns.

Methodology

The study involved preprocessing DNA microarray data and applying independent component analysis (ICA) to assess the stability of identified signatures.

Limitations

The study primarily focused on yeast data, which may not generalize to other organisms or conditions.

Participant Demographics

The study used data from wild-type yeast.

Digital Object Identifier (DOI)

10.1002/cfg.298

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication