Query-based biclustering of gene expression data using Probabilistic Relational Models
2011

ProBic: A New Method for Biclustering Gene Expression Data

Sample size: 225 publication Evidence: high

Author Information

Author(s): Zhao Hui, Cloots Lore, Van den Bulcke Tim, Wu Yan, De Smet Riet, Storms Valerie, Meysman Pieter, Engelen Kristof, Marchal Kathleen

Primary Institution: Microbial and Molecular Systems, K.U.Leuven, Leuven, Belgium

Hypothesis

Can a query-based biclustering strategy effectively identify biologically relevant gene clusters despite the presence of noisy seed genes?

Conclusion

ProBic is a robust biclustering algorithm that successfully identifies high-quality gene clusters while retaining seed genes, even in noisy conditions.

Supporting Evidence

  • ProBic outperformed other algorithms like ISA and QDB in retrieving high-quality biclusters.
  • ProBic maintained seed genes better than ISA when faced with noisy seed sets.
  • Functional and motif enrichment analyses showed that biclusters from ProBic were biologically relevant.

Takeaway

ProBic is like a smart tool that helps scientists find groups of genes that work together, even when some of the starting information is messy.

Methodology

ProBic uses a framework based on Probabilistic Relational Models to identify gene clusters by leveraging prior distributions from seed genes.

Limitations

The performance of ProBic may still be affected by the quality of the seed genes, especially in low-quality seed sets.

Digital Object Identifier (DOI)

10.1186/1471-2105-12-S1-S37

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication