SOM-based class discovery exploring the ICA-reduced features of microarray expression profiles
2004

SOM-based Class Discovery in Microarray Data

Sample size: 6221 publication 10 minutes Evidence: moderate

Author Information

Author(s): Andrei Dragomir, Seferina Mavroudi, Anastasios Bezerianos

Primary Institution: Medical School, University of Patras, Greece

Hypothesis

Can independent component analysis (ICA) and self-organizing maps (SOM) improve clustering of gene expression data?

Conclusion

The proposed method effectively clusters gene expression data while incorporating prior functional knowledge.

Supporting Evidence

  • The method integrates unsupervised and supervised learning to improve clustering.
  • Independent component analysis helps in reducing data dimensionality.
  • The approach allows for multi-labeling of gene expression profiles.

Takeaway

This study shows how to group genes based on their expression patterns using a smart computer program that learns from data.

Methodology

The study used independent component analysis (ICA) for data transformation followed by self-organizing maps (SOM) for clustering.

Potential Biases

Potential bias from reliance on existing functional classifications.

Limitations

The method may not handle all types of noise in the data and relies on the quality of prior functional knowledge.

Participant Demographics

The study focused on gene expression data from yeast.

Digital Object Identifier (DOI)

10.1002/cfg.444

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication