Skip to main content
Figure 1 | BMC Research Notes

Figure 1

From: Genome-wide matching of genes to cellular roles using guilt-by-association models derived from single sample analysis

Figure 1

Schematic for comparison of individual gene expression values to enrichment in gene sets. A) Relative, log2-transformed gene expression array data is computed for each tumor sample t (N=1,949) with tissue-matched controls. For each tumor, enrichment in each of the 4,438 gene sets described in Methods is computed. Comparison of fold-change gene expression values with enrichment scores is conducted using Spearman correlation. For a given gene, such as the hypothetical example NDUFA7, gene sets are sorted by the correlation coefficients (see Methods for detailed description). B) Enrichment in gene sets is calculated using the parametric gene set enrichment approach from Kim and Volsky (2005). Variables used in Z-score calculation are described in Methods.

Back to article page