Fig. 1 | Nature Communications

Fig. 1

From: Genomic data integration by WON-PARAFAC identifies interpretable factors for predicting drug-sensitivity in vivo

Fig. 1

The framework of WON-PARAFAC. a Individual data types represented in genes-by-cell-lines matrices with matched gene and cell line indices across the data types in which mixed sign data (i.e. gene expression and copy number) are split into two positive matrices. b Data cube formed by stacking the five matrices, where the modes span genes, cell lines, and data types. WON- PARAFAC produces three sparse factor matrixes for each of the modes: gene factor (gene-factor), cell line factor (cell-factor) and data type factor (DT-factor) matrix. c The three sparse factor matrices from WON-PARAFAC as heatmaps, with the 41st factor highlighted. d 41st factor loadings of top 15 genes (gene-factor; top left), data types (DT-factor; bottom left), and top 500 cell lines (cell-factor; right). Color of bars in the bar plot for cell-factor indicates the number of data types with CDKN2A and CDKN2B altered in each cell line

Back to article page