Figure 2.

The Score Signatures module. This figure shows the workflow for the Score Signatures module. The user supplies a gene expression data set to study (shown on left). Then, this tool will retrieve a list of (currently 18) curated signatures from the signature database, and predict their activity in each of the samples in the data set (shown on right). The software also applies hierarchical clustering, showing the patterns of pathway activation. Here, Score Signatures was applied to a panel of 19 breast cancer cell lines. The clusters show that the signatures clearly distinguish two subtypes of breast cancer. In the left cluster are the cell lines of luminal origin. The right cluster consists of all basal cell lines, with two exceptions (SKBR3 and HCC1428). The module provides a biologist-friendly interface to a complex analysis that involves statistical algorithms and curated gene expression signatures.

Chang et al. BMC Bioinformatics 2011 12:443   doi:10.1186/1471-2105-12-443
Download authors' original image