Fig. 6
From: A new strategy for Cas protein recognition based on graph neural networks and SMILES encoding
t-SNE visualizations of training sets for Model 1 and Model 2, respectively. (a) shows a tight clustering of positive Cas1 samples (red) compared to diverse negative samples (blue), validating the training set’s effectiveness. (b) highlights distinct clusters of various Cas proteins, with Cas1 proteins (outlined) forming a large, distinct group, supporting the rationality of sample selection for Model 2.
