Skip to main content
Fig. 4 | Journal of Translational Medicine

Fig. 4

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

Fig. 4

The results from validation by normal tissue RNA-seq data and patient ctcRNA-seq data. A A dot plot that shows higher correlation (r) achieved by deep learning model in the sensitivity analysis. B A dot plot that shows small MSE achieved by deep learning model in the sensitivity analysis. For A and B, the dotted line indicates the mean value. C A box violin and dot combination plot that shows smaller MSE for all 1000 semi in silico samples. D A dot plot that shows overall higher correlation by deep learning model. E A dot plot that shows the proportion of tissue contribution in 9 month follow up for a melanoma patient with brain and lung metastatic tumors. F A heatmap that depicts the complete 15 tissue results by deep learning model. G A heatmap that depicts the complete 15 tissue results by the NNLS method

Back to article page