Skip to main content
Fig. 2 | Biology Direct

Fig. 2

From: Genes sharing the protein family domain decrease the performance of classification with RNA-seq genomic signatures

Fig. 2

Spearman’s correlation distribution and violinplots of percentage of misclassified samples for genes with and without domains in CAMDA neuroblastoma dataset. On the left the red color is for the histogram-based distribution of Spearman’s correlation coefficient for a random selection of gene pairs without domains. Green color stands for Spearman’s correlation coefficient for the genes that share a PFAM domains (database built with AceView genes). Shades in the line are ranges from 100 simulations of the distribution. On the right there is violin plot of percentage of misclassified samples for 4 classifiers based on DEG with and without domains. Total number of samples in dataset was 302

Back to article page