Using pathway signatures as means of identifying similarities among microarray experiments

Luca Beltrame, Lisa Rizzetto, Raffaele Paola, Philippe Rocca-Serra, Luca Gambineri, Cristina Battaglia, Duccio Cavalieri

Research output: Contribution to journalArticlepeer-review


Widespread use of microarrays has generated large amounts of data, the interrogation of the public microarray repositories, identifying similarities between microarray experiments is now one of the major challenges. Approaches using defined group of genes, such as pathways and cellular networks (pathway analysis), have been proposed to improve the interpretation of microarray experiments. We propose a novel method to compare microarray experiments at the pathway level, this method consists of two steps: first, generate pathway signatures, a set of descriptors recapitulating the biologically meaningful pathways related to some clinical/biological variable of interest, second, use these signatures to interrogate microarray databases. We demonstrate that our approach provides more reliable results than with gene-based approaches. While gene-based approaches tend to suffer from bias generated by the analytical procedures employed, our pathway based method successfully groups together similar samples, independently of the experimental design. The results presented are potentially of great interest to improve the ability to query and compare experiments in public repositories of microarray data. As a matter of fact, this method can be used to retrieve data from public microarray databases and perform comparisons at the pathway level.

Original languageEnglish
Article numbere4128
JournalPLoS One
Issue number1
Publication statusPublished - Jan 6 2009

ASJC Scopus subject areas

  • Agricultural and Biological Sciences(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Medicine(all)


Dive into the research topics of 'Using pathway signatures as means of identifying similarities among microarray experiments'. Together they form a unique fingerprint.

Cite this