Figure 2

Procedure of vCGH training and validation. The entire dataset was split into training and testing datasets. In the training dataset, the paired GEP and CGH data were used for model parameter estimation; in the testing dataset, only the GEP data was used for vCGH prediction and the corresponding CGH data was used for validating vCGH prediction. The whole process was repeated by different splitting of training and testing datasets and the model performance was evaluated by LOOCV.