Skip to main content
Fig. 2 | Journal of Translational Medicine

Fig. 2

From: Deep neural network based tissue deconvolution of circulating tumor cell RNA

Fig. 2

Illustration of our TSG analyses. A Boxplots show that the gene MAP4K1 is primarily expressed only in the small intestine in GTEx. B Boxplots show that the gene MAP4K1 is primarily expressed only in the small intestine in normal tissues in TCGA which confirms the finding in GTEx. C A Bar chart that depicts the number of TSGs in each tissue type. Large variations in the number of TSGs can be observed. D Pie charts that the results from gene stability tests. Gray color indicates the proportion of TSGs that showed no notable change between tumor and normal. Red color indicates the proportion of TSGs that showed upregulation in tumors. Green indicates the proportion of TSGs that showed downregulation in tumors. E–G shows results from independent validation of the single tissue deep learning prediction model using TCGA normal tissue RNA-seq data. E Confusion matrix that shows overall high accuracy of the single tissue deep learning prediction model. F A polygon plot that demonstrates performance of the single tissue deep learning prediction model in F1 score, precision, and recall. G ROC curves show the high performance of the single tissue deep-learning prediction model

Back to article page