Skip to main content
Fig. 2 | BMC Cancer

Fig. 2

From: On Predicting lung cancer subtypes using ‘omic’ data from tumor and tumor-adjacent histologically-normal tissue

Fig. 2

Heatmaps for classification task a TAHNADC vs. TAHNSCC, b TumorADC vs. TumorSCC and c TAHN-TumorADC vs. TAHN-TumorSCC using the ReliefF feature selection algorithm. In the vertical axis the corresponding methylation site and gene symbol (in parenthesis) are shown. Some methylation sites do not lie in a particular gene, therefore, no symbol is provided. When multiple methylation sites are selected for the same gene, these sites should have similar methylation intensity, for it to be included. In the horizontal axis, a color-coded representation of the tissue samples is provided. Two distinct groups are observed in all three heatmaps. Cluster purity (accuracy by classification using clustering) for each task is calculated to be 1.0, 0.94 and 0.85 respectively

Back to article page